THE SYNTHESIS OF SOME 1,2-Dn!ETHYL-3-;JZYlEENZQ;ES
Aim l,2,3-TRIMirrHYL-li-ALKYLBENZENES FROM
2-ALKÏLFURAKS AND 2-METHYL-5-AIi'DfLFURANS
DISSERTATION
Presented in Partial Fulfillment of the Requirements lo: the Degree of Doctor of Philosophy in the Graduate School of The Ohio State University
By
EARL PHILLIP MOORE, JR., B. S., W. S.
The Ohio State University 1S?57
Approved by:
Adviser Department r.f Chemistrj' ACKNOWLEDGiaæNT
The author wishes to express his sincere appreciation to
Professor Cecil E. Boord for his advice and counsel*
Gratitude also is expressed to Dr. Kenneth W. Greenlee for
his continued interest and guidance.
Acknowledgement is made to Professor Boord and Dr. Greenlee
for their cooperation in making available to the writer the
equipment, materials and facilities of the American Petroleum
Institute Research Project h$»
The financial support of this work by the General Motors
Corporation, E.I. du Pont de Nemours and Company, and the American
Petroleum Institute Research Project LS is gratefully acknowledged. Table of Contents
Page
I. Introduction, 1
II. Literature Survey...... 3
A. l,2,3"Triinethylbenzene (Heraimellitene ) and l|2,3jl4“ Tetramethylbenzene (Prehnitene) from Natural Resources...... 3 B. 1,2,3“Trimethylbenzene (Hemimellitene) from De gradations of Natural Products and their Derivatives & C. 1,2,3jh-Tetramethylbenzene (Prehnitene) from De gradations of Natural Products and their Derivatives I4 D. Syntheses of 1,2,3“Trimethylbenzene (Heraimellitene) and l,2,3,U“Tetramethylbenzene (Prehnitene)...... 5 E. The Synthesis of l,2-Diraethyl-3-8thylbenzene .... 10
III. Discussion...... 11
A. General Scope . 1 1 B. The Production of 2-Alky Ifurans and 2-Methyl- 5-alkylfurans ...... Hi 1. 2-Methylfuran...... lU 2. 2-Ethylfuran...... Hi 3 « 2-n-Propylfuran...... 1 6 U. 2 j ^-Dirae thylfur an « I6 5 » 2-Methyl-^- etbylfuran...... 17 6 . 2-Methyl-^-n-propylfuran ...... 18 7. 2-Isopropylfuran ...... 19 8. 2-tert-Butylfuran...... 22
C. The Production of 2-FuryIcycloalkanes ...... 23 1. 2-Furylcyclopropane* ...... 23 2. 2-FuryIcyclopentane and 2-FuiyIcyclohexane .... 25
D« The Production of Miscellaneous 2-Substituted Furans 28 1. Furfurylidene Diacetate and Furfuryl Acetate . . . 28 2. 2-Broraofuran...... 28 3 . Tetramethylfuran...... 29
E. Production of Diels-Alder Adducts of 2-Alkyl- and 2-Methy 1-5-alkylfuran3 ndth Maleic Anhydride...... 29 1, Stereochemical Considerations...... 29 2. Preparation of A dducts ...... 32
F. The Production of 3-Alkylphthalic Anhydrides and 3-Methy 1-6-alkyIphthalic Anhydrides ...... 3 7 1. General Background...... 37 H i Table of Contents (Continucd)
2. îlydrogen Bromide M e t h o d ...... 35 3. Sulfuric Acid Method h2 , ' G. Reduction of 3-Alkyl- and 3-Methy1-6-alkyIphthalic Anliydrideo to Hydrocarbons ...... 1. General Background ...... 35 2. Copper Chromite Method ...... 3» Reduction of Anhydrides to Hydrocarbons...... 62 ij. Reduction of Diesters to Hydrocarbons...... 63 5* Reduction of Esters by Sodium and Methanol in Liquid Ammonia ...... 6^ 6. Preparation of Esters ...... 6& 7, Conclusion ...... 63
IV* Experimental ...... 6?
A* Determination of Physical Properties and Purities cf %rdrooarbon Products .6? 1* Gryoscopie Determinations , , ...... 67 2. Boiling Points ...... cd 3* Refractive Indices .65 li* Densities ...... 69
B. Special Apparatus ...... 69
C. Production of 2-Alkylfurans and 2-Methyl-3-alkyl- furans ...... 70 1* 2-Methylfuran ...... 70 2. 2-Ethylfuran...... 71 3* 2-n-Propylfuran ...... 80 ll* 2, B^-Dime thylfur an ...... 31 3 . 2-Methyl-3-ethylfursn ...... 32 6. 2-Methyl-3-n-propylfuran ...... SL 7. 2-Isopropylfuran ...... 66 8. 2-tert-Butvlfuran...... 95
D. Production of 2-Furylcycloalkanes...... 97 1. 2-Furylcyclopropane * ...... 97 2. 2-Furylcyclopentane...... 102 3 . 2-Furylcyc loh ex a n e ...... 109
E. Production of Miscellaneous 2-Substituted Furans. . .113 1. Furfurylidene Diacetate...... 113 2. Furfuryl Acetate ...... HU 3 . 2-Bromofuran 115 L. Tetramethylfuran ...... 116
iv Table of Contents (Continued)
Page
F. Production of Diels-Aider Adducts of 2-Alkyl- furane and 2-Ueth7l-5-alkylfurans 'with Maleic Anhydride 118
G. Production of 3-AlkyIphthalic Anhydrides and 3-Methyl-6-alkylphthaHc Anhydrides...... 126 1. Hydrogen Bromide Method ...... « 126 2, Sulfuric Acid Method* 127 3* Other Aromatization Experiments ...... 132
H. Reduction of 3-Alkyl- and 3-Methyl-6-alkylphthalio Anhydrides to Hydrocarbons* •••••••••••• 133 1. Preparation of Phthalate Esters .•••*.•« 133 2« Copper Ghrmiite Method* **** *** *** 136 3* Reaction of Phthallc Anhydrides to Hydrocarbons 139 1;* Reduction of Phthalate Esters to Hydrocarbons * 1U2
V. Summary...... IL?
VI. Autobiography...... Index to Tables
Table No. Page
I Syntheses of 1,2,3“Trimethylbenzene (Hemimellitene) 6
II Syntheses of 1^ 2, 3,li-Tetramethylbenzene ( Prehmitene ) 8
III Physical Properties of Hydrocarbons 9
IV Classes of Substituted Furan-Maleic Anhydride Adducts 35
V Hydrogen Bromide Aromatization Experiments ijl
VI Sulfuric Acid Aromatization Experiments hS
VII Aromatization Experiments with Various Substituted Furan-Maleic Anhydride Adducts {96% Sulfuric Acid at -15* to -10®) L8
VIII Other Aromatization Experiments 52
IX Elemental Analyses of New Diels-Alder Adducts 1 2 k
X Melting Points of Diels-Alder Adducts 125
XI Alkylbenzoic Acids from Aromatization Experimentsl25
XII Elemental Analyses of New Phthalic Anhydrides I30
X m Melting Points of Phthalic Anhydrides I3I
XIV Physical Properties of phthalate Esters 136
XV Elanental Analyses of Phthalate Esters I36
XVI Elemental Analyses of Diols 21x3
XVII Physical Properties of Diols lijli
XVIII Elemental Analyses of New Hydrocarbons 2lx$
XIX Physical Properties of hydrocarbons 1L6 vi INTRODUCTION
In the latter part of the nineteenth century, much interest in
pvorslble synthetic routes to 1,2,3"trimethylbenzene (hemiaellitene )
and 1,2,3,h-tetramethylbenzene (prehnitene) was actuated by absorp
tion spectra studies of the isomeric tri- and tetramethylbenzenes.
The above tiro aromatic structures were the last of these isomers to
succumb to the efforts of early investigators and indeed, to the
present time, continued interest in synthetic approaches to heni-
melütene and prehnitene have occupied investigators in the field
of organic chemistry.
It was not until the advent of World War II, when S.F. Birch and
his co-TTorkers began studies of the ultra-violet spectra of the iso meric dimethylethylbenzenes, that another l,2-dimethyl-3-alkylbenzene was prepared; their work was published in 191^9. Since then the
synthesis of no other member of this class of vicinally substituted baizenes or of the 1,2,3-trimethyl-f4-alkylbenzene class has been re ported in the chemical literature. From information obtained from this source regarding the transformation of Diels-Alder adducts of substituted furans and maleic anhydride into aromatic anhydrides through the agency of acid catalysts, possible isomer-free routes to the above aromatic types materialized.
The main effort of this research program therefore was in vested in determining the best method of aromatizing a number of
2—alkyl- and 2-methyl-^—alkylfuran-maleic anhydride addends and
- 1- -2" converting the resulting 3-alkyl- and 3-methyl-6-alkylphthalic anhydrides into hydrocarbons, I'bcistlng procedures for the preparation of the required furan precursors were tailored to enlarged-scale work and new ones were devised when necessary. Also included In this work were nhe preparations of three 2-furyIcycloalkanes (two of them new} and several miscellaneous 2-3ubstituted furans, and the application of the best aromatization method to tlieir adducts*
The physical properties of those hydrocarbons which were syn thesized were determined in order to aid in future identifications of them. LITERATl'RE SIjJÎV?T
1.2.3-Triinethylbf^nsftno (ngri-Lv.eUi tone ) .ind 1.2,3.^~TetrôJicfchyl-
benzene ( Prehnitene ) from Natural ilesources
Although both hemimelliter.t; and prehnitene were first prepared
in the l800^s, it was not suspected for a good many years afterward
that these compounds occurred in the gasoline or naphtha fraction of
petroleum and in the low temperature oils removed from coal tar
during its destructive distillation. The reason for this oversight
can be attributed to the presence of only very small amounts of these
vicinally substituted aromatics relative to the quantities of close-
boiling isomers, for in X 9 2 h Kruber^ reported that very carefully
repeated fractionations of low temperature tar oils ( 190300"F. ) reve"al6d the presence of a compound which exhibited the properties of hemimellitene* Several years later Birch®, in the first of his many investigations of petroleum and coal tar products, isolated hemimellitene from Iranian petroleum, Mair and Schicktanz^ later found it in Oklahoma petroleum, Sy controlled sulfonation of coal tar naphtha, Eirch^ found that he could concentrate this isomer. It was not until I9UI that prehnitene was isolated from petroleum® and until 19^1 vi-hen it was removed froa coal tar light oil®.
1 0, Kruber, Ber,, IOO8 Tl92i: ), 2 S.F. Eirch, J, Chem, Soc,, 1926. 3 B,J. Mair and S,T, Schicktann, J. Res, Natl, Bur. Stand., 11, 665 (1933). 4 S.F. Birch, Brit. Patent 565,336 (19l7). 6 B.J. Mair, J, Res. Natl. Bur. Stand., 2%, 3U3 (19ll). 6 0. Kruber and A, Raeithel, Brennctoff-Chemie, 2h3 (1951), - h -
1,2, j-Trlmethylbenzone (Hcninielli t.ene ) I’rcim Lf^çrüdationg of Natural
Products and their Derivatives
The first report of the presence of hemimellitene amo;.^ tlie
residues of a degraded natural product was made cy Ruzicka"^ in 1931;
he found that a tricarboxylic acid (obtained when abietic
acid was subjected to permanganate oxidationJ gave small amounts of
this hydrocarbon when dehydrated— dehydrogenated with selenium. He
later isolated a little hemimellitene and several ether aromatic com
pounds when 1,1,2-trimethylcycloheptane and 1, l,Li-trimethylcycloheptane
(eucarvanc )** were subjected to similar degrading conditions, as did
Pines when he dehydrogenated pinane® and d-limonene^° over aluminum
oxide. Investigators also have reported that very small amounts of the aromatic were obtained when allocimene^^ and plinol^^ were pyrolysed in the absence of a catalyst*
1.2.3.L-Tetramethylbenzene (Prehnitene) from Degradations of Natural
Products and their Derivatives
During his many investigations of the structures of triterpenes and triterpenoids Puzicka repeatedly discovered prehnitene as a minor product of dehydrogenations with selenium- and platinum-cn-aluminum
7 L. Ruzicka, et al., Helv, Chim. Acta, iL, 5 h S (1931). e L. Ruzicka, si ai*, Helv, Chim. Acta, Ig, li2i: (1936). b H. Pines, J, Am, Chem. Soc*, JO, ^33 (19Û8). 10 H. Pines, J, Am. Chem. Soc., ^ L8?2 (19S2). 11 B.C. Parker and L,A, Goldblatt, J. Am. C h & s , Soc., J2, 21^1 (1950). T. Sebe and T. Naito, J, Taiwan Fharm. Assoc., ^ No, 1, 23 (19^0). cxicie catalysts* Betu3J.n, siaresinolic acid^^j betulinol^'*; ht;dera-
jenin, oleanolic acid, sumarcsinollc acid^®; and c L and ^ aroyrin^^
yielded verj' small quantities of prehnitene together with other sub
stances. In addition, Nol.ler dehydropenated echinocystic acid^'* and
Japanese investigators pyrolyzed ursolic acid-^® arid skimrrlol^®;
however, these investigators obtained many other products in addition
to the above hydrocarbon.
Syntheses of 1. 2.3"Trimethy lb en 2 an e ( H emimelli t en e ) and 1.2.3^1t"Tetra-
methyIbenzene (Prehnitene)
A complete survey of different methods of synthesizing these
compounds has been included in Table I and Table II,
13 L. Suzicka, et Helv, chim. Acta, Ig, lL9c (1932), 14 L. Ruzicka, et Helv. Chim. Acta, 17, U26 (193U). 15 Ibid. p. L. Ruzicka, Helv, Chim. Acta, 20, 791 (1937). 17 C.H. Holier, J. Am. Chem, Soc,, 1^82 (193li). la K. Huzii and S. Osumi, J, Pharm. Soc, Japan, $9. 711 (1939), IB K. Takeda and S. Yoshika, J, Pharm, Soc. Japan, 61, $06 (I9ll), -6—
Table I
Syntheses of 1,2,j-Trimethylbenzene (hgrdmellltene }
Method of Synthesis (Final Step) Over-all Yield Reference
1} Calcium 3ilj,5"triinethylbenzoate Not reported 20 heated with lime
2} Calcium 2,3,ij-trimethylbensoate Not reported 21 heated with lime
3 ) 2-Bromo-l,3-dimethylbenzenej Satisfactory 22 methyl iodide, sodium, heated in ether li) 3-Iodo-l, 2-dlmethylbenzene, methyl Very low 23 iodide, sodium, heated in ether
5) 2,3-Dimethylbenzyl chloride (pre 26% 2h pared from benzyl chloride by successive Tiffeneau rearrange ments) reduced by sodium amalgam
6 ) 1,2,3-Trimethylcyclohezene fraction, 2k% 25 b.p. Ilj6-1^6* (obtained from piperylene-crotonaldehyde adduct by catalytic reduction and dehydration) dehydrogenated over chromia-alumina catalyst
7 ) 1,2, 3-Trime thyl-l-cyc loh exene ( ob 62% 26 tained by dehydration of product of CH^MgBr and 2,6-dimethylcyclo- hexanone) dehydrogenated over alumi num oxide
20 0. Jacobsen, Ber». l5. 1657 TÎ882). 21 0. Jacobsen, Ber., 1£, 1215 (1886). 22 0* Jacobsen, Ber., 20, 901; (1887). 23 K. von Auwers, Ann., Iil9. 116 (1919). 24 L.I. Smith and L.J. Spillane, J. Am. Chem. Soc., 62, 2639 (19hO)* 20 T.B. Tom and G.E. Boord, I0?th Meeting, American Chemical Society, April 19Uit, Cleveland, Ohio. 26 J.E. Nickels and W. Heintzelman, J. Org. Chem., I2 , 11U2 (1950). -7-
Table I (Continued)
Method of Synthesis (Final Step) Over-all Yield Reference
Ô) 2,3-Dimethylbenzyl chloride (pre 22-2h% 27 pared by chlorométhylation of o-xylene) reduced by LiH-LiAlH^
9) Chloromethylated 1,3-dimethyl- 28 9-tert-butylbenzene reduced by LiH-LiAlH^ and tert-butyl group removed
10) Piperylene-crotonaldehyde adduct 29 heated with phosphorus pentoxide
11) 2,3-Dimethylbenzyldimethylamine 30 (obtained by successive Stevens rearrangements, starting with benzyltrimethylammonium iodide) reduced with sodium amalgam
12) 1,1,2,3-Tetramethylcyclohexane ho% 31 dehydrogenated over platinum-on- aluminum oxide
13) 1,1,3-Triinethylcycloh exane de 10^ 32 hydrogenated over platinum-on- aluatlnum oxide in presence of secondaiy-butyl chloride
111) 2,2,li-Trimethylcyclohexane- Not reported 33 carboxyllc acid dehydrogenated- decarboxylated over nickelous oxide
Yield is for final step only.
27 C.D. Shacklett and H*A. Smith, J. Am. Chem. Soc., 766 (19^1). 28 M.J. Schlatter, J. Am. Chem. Soc., ^6, U9S2 (1951»)• 29 A.A. Petrov, et Zhur. Obshchei. Khim., 21^, 298 (195Ü). 30 Tr.R. Bras en and C.H. Hauser, Organic Syntheses. Vol. 3h, John Wiley and Sons, Inc., New York, N.Y., 19^1», p. 56ë 31 V.H. Ipatieff and H. Pines, U. S. Patent 2,li35,l»33 (19ii8). 32 H. Pines, J. Am. Chem. Soc., 6226 (1953), 33 H. Pines, Bull. soc. chim. France, 1951. 259* -e-
Table II
Syntheses of 1,2,3,!j-Tetramethylben%ene (prehnitene)
Method of Synthesis (Final Sten) Over-all Yield Reference
1) Pseudocumene, methyl iodide, Very low 3h aluminum chloride, reacted in carbon disulfide
2) 2,h-Dibromo-l,3-dimethylbenzene, Mot reported 3? methyl iodide, sodium, heated in ether
3) 3-Bromo-l,2,ü-trimethyIben2 en e, Very low 36 methyl iodide, sodium, heated in ether k ) Pentamethylbenzene, subjected 19% 37 to Jacobsen rearrangement
5) Durene and isodurene, subjected hl.U% 37 to Jacobsen rearrangement
6) 1,2,3i^“Tetramethylcyclohexane 3 1 % ^ 38 dehydrogenated over selenium
7) Xylenes, b.p. 138-1^5°, methyl Very low 39 iodide, aluminum chloride, re acted in carbon disulfide
8) 5“Ethylpseudocumene and ethyl- 10% ^ Uo mesitylene, subjected to Jacobsen rearrangement
Yield is for final step only.
34 A. Claus and C. FoecJcing, Ber., 20, j097 (1887). 35 0. Jacobsen, Ber., 21, 2823 (iSdF). 36 Ibid. p. 2827. 37 L.I. Smith, Organic Reactions. Vol. I, Chapter I, John Wiley and Sons, Inc., New York, N.Y., 191^2, p. 371. 30 D.I. Mitchell and C.S. Marvel, J. Am. Chem. Soc., 55» 1|276 (1933). 39 J. Savard and R. Hosognt, Rev, Fac. Sci., Univ. Istanbul, 3, 27 (1937). 40 L.I. Smith and II.A. Kiess, J. Am. Chan. Soc., 989 (1939). -9-
Table II (Continued)
Method of Synthesis (Final Step) Over-all Reference
9) 2,3,L-Trlmethylbenzyl chloride 71-7$^ 27 (prepared by chlorométhylation of hemimellitene) reduced by LiH-LiAlH^
1 0 ) Dichlo!*omethylation product of 19-20$ 27 m- or £-:cylene reduced by . Ü H - U A I H 4
11) 3,6-Dime thy 1-v^xylyl ene brond.de $6$ lil reduced by LiAlH^
Table III
Physical Properties cf Hydrocarbons^^
Hydrocarbon b.p. ngo d20 f.p. 4
l,2,3"Trimethylbenzene 176.08U" 1*^1393 0.69138 -2$.37$*
1,2-Dime thyl-3-ethyl- 193.91* l.^H? 0.6921 -U9.$“ benzene
1,2,3,14-Tetramethyl- 20$.OL" l.$2G3 0.90$2 - 6.2$* benzene
41 E. Buchta and G. Loew, Ann., $97. 123 (19^6). 42 F.D# Rossini, e t Selected Values of Properties of Hydrocarbons ^ d Related Compounds, American Petroleum Institute Research Project m r The Carnegie Press, Carnegie Institute of Technology, Pittsburg^:, Pa., 19^2, Tables liia. -10-
Tlie Synthesis of 1, P-Dimethyl-^-othylbenzene
The synthesis of this compound has been reported but once in the
chemical literature‘s^. 1-Nj.tro-l^2-dimethylbenzene was reduced by
catalytic hydrogenation to 3-anino-l_, 2-dime thylbenzene, which in turn
was diauotized and converted to the 3-bromo derivative by a Sandmeyer
procedure. The Grignard reagent prepared from this compound reacted
irLth acetonitrile to give 2 ,3-dimethylphenj’-l methyl ketone (2,3-di-
methylacetophenone), which was reduced to the desired hydrocarbon by
a Glemraensen procedure. An over-all yield of less than was
realized.
43 S.F. Eirch, et J. Am. Chem. Soc., 1362 (19li9). DISCUSSION
General Scope
A s haô already been poLnted out, synthetic routes to only two of
the numerous possible 1, 2 -dime thy 1-3-alkyl- and 1, 2, 3“triinethyl-
ii-alkylbenzenes have been of interest to investigators in the field
of organic chemistry: 1,2,3"trLmethylbenzene (hemimellitene)and
l,2 ,3^h-tetramethylben2 ene (prehnitene). One other compound (1,2-dl- methyl-3-sthylbenzene), a member of the first of the above two
classes, has been prepared in very low yield. No evidence could be
found that any investigation had been carried out in order to deter mine the utility of a method of aromatic hydrocarbon synthesis en tailing the aromatization of Diels-Alder adducts of substituted furans and maleic anhydride and the subsequent reduction of the phthalic anhydrides produced. Such an approach should permit isomer- free syntheses of many valuable compounds and obviate the use of elaborate purification schemes sometimes necessary to separate them from close-boiling isomers. Of exceptional importance would be the application of 2-alky If urans and 2-me thy 1-5 -alky If urans to the con struction of aromatics of the classes mentioned above, as illustrated by the following general scheme:
-11- -12-
=c»
trhere R = -CH3 or alkyl
and R = -H or “CH3
Based upon two reports in the chemical literature that certain adducts of 2-3ubstituted furans and maleic anhydride had been suc cessfully aromatized, a cursory preliminary investigation was made with several adducts in order to ascertain if yields of the aromatic anhydrides obtained by either of these methods would be high enough to justify further work.
Because of the gratifying results of this investigation a research program was undertaken to determine the best aromatization method and conditions allowing optimum yields, to work out a satisfactory re duction scheme, and finally to apply these findings to the synthesis of a number of 1,2-dimethy 1-3-alkyIbenzenes and 1,2,3-trimethyl-
U-alkylbenzenes, several on an enlarged scale.
Three furans— 2,5-diaethylfuran, 2-methylfuran and furan itself— were commercially available for this program and from the latter two almost all of the others were prepared. In several instances, mcdi- ficationo of known procedures were found to be necessary when molar quantities of reactants were increased, and in one case a new method of reducing an olefinic side chain without affecting the furan nucleus -13- had to be devised; it vras fui'ther applied in the syntheses of three oth;,r furans (two of them— mentioned below— were new 2-furylcyclo- alkanes},
In order to expedite the formation of large qucintities of adduct, and to Insure the formation of finely divided material in certain cases, several modifications of the usual procedure for Diolc-Alder condensations of furans with maleic anhydride were made.
In addition to the furans necessary for the preparation of
1,2-dimethyl-3-alkyl- and 1,2,3-trimethyl-h-alkylbenaenes, three
2-furylcycloalkanes (two of them new) and a number of miscellaneous
2-substituted furans were synthesized and the proven aromatization procedure was carried out with their adducts.
A secondary purpose served by this research was the determination of the physical properties of the hydrocarbons synthesized in order to aid in their future identifications. -1 h-
Thc rroduction of ?-Alkylf urans and 2-M
2-l'ethyll'uran.— This furan ivas obtained ccmnercially from the
1.1. du Pono de Piaours and Co., T/ilmlngton, Cclav.are. The ccmmeroial
product first v.as distilled carefully through a packed colrnnn of about
fifteen-plate efficiency, but the obvious presence of a close-boiling
impurity gave only distillate of quite low refractive index. A slow
refractionation of this material, using a packed column of fifty-
plate efficiency, yielded good 2-methylfuran (b.p.
n^° 1.1:332 ). A second compound distilled at 66-66*1“ and showed a n^®
of l.Ul5 0 -l*l,'l5 5 j it amounted to about 13% of the 3.2 kg. of material
fractionated. Positive identification of the structure was not made;
however, its infra-red spectrum indicated that it might be 2-methyl-
2,3-dihydrofursn, e compound unreported in the literature. Other
2-methyl-dihydrofurans and methylenedihydrofurans have quite different
physical properties,
2-Ethylfuran .— Several possible means of preparing this compound were examined. The first of these was essentially the one described
by Paul^> His procedure consisted of: 1) a Grignard reaction between
furfural and méthylmagnésium iodide to give 2-furylmethylcarbinol,
2 ) dehydration of the alcohol by passage of it through a column packed
1 R. Paul, Compt. rend., 200, ll:8 l (19351» -15~
-.rith anhydrous aluminum c x l d e and maintained at , 3) hydro
genation of the resulting P-vinylfuran in the presence of a platinum
catalyst. No procedural details or ji,elds of products v/ere given.
Although 2-furylir.ethylcarbinol could be prepared in respectable
yields by reacting furfural and méthylmagnésium iodide accordir^g to
the method of Yurev an: Gragerov^, the remainder of Paul*s procedure
proved tc be unsatisfactory' for the preparation of 2-ethylfurar. for
several reasons: Ij the dehydration step yielded a considerable
amount of polymer and 2 ) the hydrogenation of 2-vlnylfuran over the
platinum catalyst preceded very slowlyj also, 2-cthyltetrahydrofuran
T.'8S formed in addition to 2-ethylfuran. A palladiurn-on-charccal
catalyst was substituted for the platinum one, with no better result.
In the present research the weaknesses in this method eventually
were overcome. A smooth dehydration was found to take place if the
alcohol and a trace of iodine (contained in a Claisen flask assembled
for vacuum distillation} were heated h y a stoam bath and pressure was
reduced in the system sufficiently to remove the olefin as it was
formed. Yields of olefir; up to 6$X were obtained in this manner,
2-Vinylfuran (it was discovered later in this research program) can
be reduced to 2-othylfuran of exceptional purity in 96^ yield by
sodium in liquid ammonia.
The second and accepted procedure was one which could be carried
out in two simple steps and on an enlarged scale. PUran was
2 Y.K. Yurev and I.P. Iragerov, Zhur. Obshchei Khim. 19, 72h (19L9}: C.A., 1092 (1950). -16-
acetylated by acetic anhydride in the presence of a boron trifluoride-
diethyl ether catalyst, by the method of Heid and Levine^, to give
2-fury1 methyl ketone in yields; when carefully distilled, it
solidified at room temperature into white crystalline material. Re
duction of the ketone to the desired furan was accomplished in 7T-80%
yields by a modified Wolff-Kishner procedure^.
One other acétylation procedure was tried, utilizing boron tri-
fluoride-acetic acid® as the catalyst. It gave yields of ketone simi
lar to those obtained when boron trifluoride-diethyl ether was used.
2-n-Propylfuran.— This furan was prepared by the same procedure
as that used to prepare 2-ethylfuran. Propionylation of furan with
propionic anhydride in the presence of boron trifluoride-diethyl ether
(according to Heid and Levine^) took place rapidly and smoothly to
give 6o-66j6 yields of ethyl 2-furyl ketone. Reduction of the ketone
by the modified Wolff-Kishner method^ to 2-n-propylfuran took place
in 8 O 8 0 yields.
2,^-Dimethylfuran. — As in the case of 2-methylfuran, this com
pound was obtained coumercially (Carbide and Carbon Chemicals Coarp. ).
Unfortunately, the same difficulty which was encountered in the
purification of 2-methylfhran also was experienced here.
3 J.V. Heid and R. Levine, J. Org. Chem., 1 ^ 1^09 (191^8). 4 l8th Annual Report, American Petroleum Institute Research Project 1956, p. 59. 6 R. Levine, J.V. Heid, and M.W. Farrar, J. Am. Chan. Soc., 71, 1207 (19U9). -17-
An initial distillation of the crude product through a column of
about 15-plate efficiency did not effect the purification of 2,5-di-
methylfuran; the fractions collected distilled at lower temperatures
and exhibited considerably lower refractive indices than those report*
ed® for the pure compound (b.p. 9U.I/760 mm., n§° l.WilO). A second
distillation (of the combined fractions) through a column of 50-
plate efficiency yielded furan having the correct boiling point but
low refractive index (n^° l,i>hOO maximum). An impurity, which was
not identified, distilled at 91»5*/7U8 mm. (ng° l.li2UU-l.l*2l*8). It
amounted to about 8% of the 3.2 kg. of material collected. Fortunate
ly, the impurity idiich remained in the 2,5-dimethylfuran did not
appreciably affect the yield of maleic anhydride adduct, »faich was
excellent.
2-Methyl-5-ethylfuran— The preparation of 2-methyl-5 - e thylfur an,
as in the preparations of 2-ethyl- and 2-n-propyl furan, was carried out according to a procedure which embodied an acylation step and a reduction step*
Only one acétylation procedure reported in the literature has claimed more than a token yield of 5-methyl-2-furyl methyl ketone, that of Farrar and Levine’, in irtrlch a boron tidfluoride-diethyl ether catalyst was used to catalyze the reaction of 2-methylfuran and acetic anhydride. These authors worked with only one-half molar
6 18th Annual Report, American Petroleum Institute Research Project L5, 1956, Cumulative Table. 7 M.W. Farrar and R* Levine, J. An. Chem. Soc., J2, 3695 (1950). -16»
quantities of reactants, obtaining the desired ketone in h2% yield.
When this procedure was scaled up by a factor of twelve much polymer
was encountered and yields of ketone for a number of runs never ex- ^
ceeded !$%• A good deal of the difficulty was overcome by applying
more efficient initial cooling in order to temper the vigor of the re
action, and also by employing a rapid vacuum-stripping operation prior
to final distillation in order to remove the ketone from the initially
formed polymer and minimize further loss. These two precautions in
creased the yield to a maximum of 35^* The attempted removal of
ketone from the reaction mixture by steam distillation only accelera
ted further polymerization. Fortunately, the over-all yield was fair,
as 8Of conversions of the ketone to the furan were realized in the
modified Wolff-Kishner reduction step*.
2-Methyl-^-p-propylfuran. — Farrar and Iæ vines reported the prep
aration of $-metbyl-2»furyl ethyl ketone in fair yield through the
boron trifluoride-diethyl ether catalyzed reaction of 2-methylfuran
and propionic anhydride. In the above investigators' procedure
(carried out on a one-half mole scale) the catalyst maa added «11 at once to the reactants at room temperature instead of 0* as in the previous acylations discussed. In the presently described research tarried out on a 3-mole scale) it was found that a better yield of ketone could be obtained if the vigor of the reaction was moderated somewhat by placing the reaction vessel in a bath containing warm e M.W. Farrar and R. Levine, J. Am. Chem. Soc., 22, 369$ (1950). -19-
irater and also by employing a vacuum-stripping operation preliminary
to final distillation. A maxinnun yield was realized. 9-Methyl-
2-fury 1 ethyl ketone was reduced to 2-methyl-5”n-proiîylfuran in 8 0 *82$
yields by the method utilized for the previous reductions.
2-Isopropylfuran»— Extended efforts were made to prepare suffi
cient quantities of 2-isopropylfuran by the two methods which had been
reported in the chemical literature, but without success. The first
procedure which was tried, that of Gilman and Calloway®, was dis
carded after repeated attempts were made to carry out the initial
step. The second procedure, that of Reichstein^®, was shown to be unsatisfactory except for the preparation of a very small quantity of material. The shortcomings of these proctsdures and efforts to over come them in the presently described research are related in the following paragraphs.
1. Gilman and Calloway reportedly effected the preparation of
2-isopropylfuran in three steps: a) equlmolar quantities of methyl
2-furoate and isopropyl chloride (or n-propyl chloride) were added to a suspension of anhydrous aluminum chloride (one-two moles/mole re actants) in carbon disulfide (at 0*), and the product was worked up for methyl $-isopropyl-2-foroate, b) this ester was hydrolyzed then by base to ^-isopropyl-2-furoic acid and finally, c) the acid was decarboxylated by copper-bronze in quinoline at 210'. e H. Gilman and N. Calloway, j. Am. Chem. So*., ^ Ul97 (1933). 10 T. Eei<*8tein, Helv. Chim. Acta, 1^> (1932). -20-
Tflfhen the first step was attempted by the present author, using
one mole of aluminum chloride per mole of reactants, only evil-smell
ing orange-red oils and much polymer were obtained. Consequently, a
number of experiments were carried out in an effort to eliminate these
undesirable products. The mole ratio of aluminum chloride to re
actants was varied up to 2/1; excess isopropyl chloride (or n-propyl
chloride) was used; aluminum chloride was added to the reactants in
an inverse addition procedure; low reaction temperatures were tried.
All experiments were unsuccessful; only colored oils and polymer were
obtained. It was concluded from the results of these experiments that
carbon disulfide had been taking part in the reaction (Gilman and
Calloway® previously had reported that such a participation took place
when the Friedel-Crafts reaction of methyl 2-furoate and a methyl or
ethyl halide was carried out in carbon disulfide). Therefore, two
other solvents— nitrtanethane and sym-tetrachloroethane— were tried.
Isopropyl chloride and methyl 2-furoate reacted in nltromethane to
give only polymer, but a 20^ yield of methyl $-isopropyl-2-furcate was obtained when the reaction was carried out in the chlorinated hydrocarbon. No farther effort was made to use this last procedure; also, the ester w M c h was prepared was not converted to the fur an.
2. Reichstein's procedure also was a three-step one: a) di me thy 1-2-furyIcarbinol was prepared in good yield from 2-fuiyl methyl ketone and méthylmagnésium iodide, b) the alcohol was dehydrated (in low yield) by refluxing a mixture of It and an equivalent amount of -21-
acetic anhydride and c) the 2-isopropenylfuran (2 g. ) was reduced by
hydrogen to 2-isopropylfuran, using a platinum oxide catalyst (one-
half hour required).
Bachmann and Heisey^^ later prepared 2-isopropenylfuran according
to a more desirable method. Dime thy 1-2-fury Icarbinol was prepared
from' methyl 2-furoate and méthylmagnésium iodide, and the alcohol was
dehydrated by refluxing a mixture of it, an equivalent amount of
acetic anhydride, and one-half of an equivalent of anhydrous sodium
acetate* Since methyl 2-furoate is more easily prepared than 2-fuiyl
methyl ketone and the use of sodium acetate gave a higher yield of
product in ttie dehydration step, this method was used in the presently
described research in preference to Reichstein’s for the preparation
of most of üie furylalefiii. required.
The present author found Reichstein^s reduction method to be too
slow to be practical at room temperature or below and uncontrollable
at h i ^ e r temperatures (giving rise to much 2-isopropyltetrahydro-
furan). A search of the literature revealed that Railings and Smith^* were able to reduce (selectively) unsaturated side chains attached to the fur an nucleus by Iqrdrogen in the presence of a palladized stront ium carbonate catalyst (all of the compounds reduced were unsaturated acids or alcohols ). However, when this catalyst was tried for the reduction of 2-isopropenylfuran, the same difficulties were encount ered as with platinum oxide. Other catalysts tried were palladium-
11 G.3. Bachmann and L.V. Heisey, J. Am. Chai. Soc., n , 1985 (19h9), 12 R.J. Railings and J.C. Smith, J. Chem. Soc., 1953.“^^ . -22-
on-charcoal and platinuia-on-asbestoa. Both were found to be un
satisfactory for this purpose at room temperature or below.
Birch^^, in IShS, reported that primary, secondary, and tertiary
benzyl-type alcohols could be reduced smoothly to the hydrocarbons by
sodium and alcohol in liquid ammonia; however, reduction of a furan
analog, 2-furylisopropyIcarbinol, gave only 10% 2-isobutylfuran. The present author nevertheless tried the method on dimethy1-2-fury1-
carbinol, but obtained no 2-laopropylfuran^ half of the alcohol was recovered unchanged. Although phenyl substituted olefins (styrenes) have long been known^^>^® to react with sodium in liquid ammonia to produce the corresponding alkylbenaenes, no example of a reduction of a furyl substituted olefin to an aUcylfuran by this method could be found in the chemical literature; it was decided therefore, to try this method on 2-lsopropenylfuran« When one mole of the furan was subjected to the action of 2 moles of sodium in liquid ammonia,
2-isopropylfuran was obtained in 90% yield. Reductions of larger quantities were carried out with the same excellent results*
2-tert-Butylfuran— No difficulty was encountered in the prepara tion of this furan. Gilman and Calloway® reported its synthesis in good yield by the same procedure used to prepare 2-isopropylfuran.
Although methyl ^-lsopropyl-2-furoate could not be obtained by iso- propylation of methyl 2-furoate (p. 20), tert-butylation of this
13 A.J. Birch, J. Chem. Soc., 19h^. 809* 14 P. Lebeau and M. Picon, Compt. rend..1$7. 223 (1913). 15 C.B. Wooster and J.F. Ryan, J. Am. Chem. Soc., 1133 (193U). -23-
ester gave methyl 5-tert-batyl-2-furoate in yields as high as
80%. Subsequent alkaline hydrolysis of the alkylated ester gave a
quantitative yield of the acid, which was decarboxylated by copper-
bronze in quinoline at 210* in 70-80^ yields.
fil'l of the furans prepared were systematically characterized in
the course of this research. Not only were their Diels-Alder adducts
with maleic anhydride prepared (with the exception of 2-tert-butyl-
furan), but their boiling points at specific pressures and their re
fractive indices were recorded. These will be given in the experi mental chapter and compared with those recorded in the literature.
Intermediate ketones and 2-furylolefins also have been characterized by appropriate means.
The Production of 2-Farylcycloalkanes
2-Farylcyclonropane « — The preparations of two 2-fury 1- sub stituted cycloalkanes have been reported in the literature and in each case it has been a 2-furylcyclopropane. In 1929 1 Kishner^® prepared 1- ( 2-furyl )-2-methylcyolopropane by heating furfuryl-
Ideneacetone with a two-fold excess of 90% hydrazine hydrate in absolute ethanol and decomposing the formed 3-methyl-$-(2-furyl)- pyrazoline at 2CX)* with potassium hydroxide in the presence of platinized porous plate. In lÿl*l, Shulken and Daiber^*^ synthesized xe N. Kishner, Bull. soc. dhim. France, 767 (1929). X7 N.I. Shnikin and 7.V. Daiber, Bull. acad. soi. U.S.S.R., Classe sci. chia., 19L1. 121» -2h“
2-furylcyclopropane by the same procedure, using 3~(2-furyl )acrolein
as the unsaturaved carbonyl compound; no yields or physical properties
of products were given. Since 3~(2-furyl)acrolein can be readily ob
tained in high yield by the method of Hinz^** (by condensing furfural
and acetaldehyde at low temperatures in the presence of a small amount
of sodium hydroxide ), it was decided to include 2-furylcyclopropane
among two other 2 - fury Icy cloalkane s to be studied.
The reaction of the aldehyde and 9^% hydrazine proceded with
vigor, but when the product was vacuum distilled, apparent decompo
sition took place throughout the distillation. Even though large
excesses of hydrazine were employed in order to insure complete con
version of aldœyde to pyrazoline, a considerable residue of the
aldazine always remained. Impure ^-(2-furyl)pyrazoline was subjected
to repeated vacirnm distillations at low pressures, but each time some
aldazine was formed and hydrazine was evolved, thus resulting in a
20-25^ loss of product.
The platinum-catalyzed decomposition of 5“(2-furyl)pyrazoline
yielded, as a rule, a mixture containing 2-furylcyclopropane and
2-propenylfuran in a ratio of about 3/5* These furans could not be
separated cleanly by distillation. However, when the mixture was
treated with a solution of sodium in liquid ammonia, the 2-propenyl— furan was reduced to 2-n-propylfuran; the two components then were
18 A. Hinz, G. JÉsyer and G. Schucking, Ber., ?6b. 676 (19^3), separable by distillation dne to ttie v/ider difference in boiling
points.
2-tiiryIc:'/cIonentane and 2-Furyleycloiiexrme.— These new compounds
were prepared by the came procedure, v.bich incorporated tvr'o nerr re
actions of furan compounds.
Syntheses of the above cycloalltanes could not be accomplished by
the method of Gilman and CaHoway® (which had proven to be quite use
ful for the preparation of 2-tert-butylfursn but not for 2-isopropyl
furan}. Attempted alkylations of methyl 2-furoate with cj’rlopentyl
or cyclohexyl halides (bromide or chloride) in the presence of alumi
num chloride and in several solvents invariably gave much polymer and
approximately half of the methyl 2-furoate was recovered unchanged.
A second possible route which was considered would have required
the initial preparation of the 2-furyl Grignard reagent, which then
could have been allowed to react with the proper cyclic ketones to
give the 2-furylcycloalkanols. Subsequent dehydration of these
alcohols and reduction of the olefins formed would have furnished the desired 2-furylcycloalkanes.
However, the 2-bromofuran^^'^° or 2-iodofuran2i required for the preparation of the above reagent can not be obtained as readily by
A*F* Shepard, N.R. Winslow and J.R. Johnson, J. Am. Chem. Soc., ^ 2082 (1930). 20 R.E. Lutz and J.M. Smith, J. Am. Chem. Soc., 62, H L 8 (I9J4I).
H. Gilman, H.E. Mallory and G.F. Wright, J* Am, Chem. Soc., 733 (1932). -26-
existing procedures as another reagent, 2-fury111thium, which can be
formed in good yields upon metallation of furan by both phenyllithixun^®
and n-butylllthium^ ^. No examples of reactions of this organoasetallic
with aldehydes or ketones have been reported, although it has given a
fair yield of 2-furoic acid when treated with carbon dioxide**.
A procedure therefore was chosen which could be carried out in
four simple steps; 1 ) metallation of furan by g-butyllithium;2 ) re
action of the 2-furyllithium with cyclopentanone (or cyclohexanone),
3 ) dehydration of the resulting alcohol under mild conditions (non
acid) and U) reduction of the cycloalkenyl moiety with sodium and
liquid ammonia*
n-Butyllithium, prepared according to the directions of Gilman
and co-workers***, was not filtered before use because a loss of about
10^ of the organometallic would have resulted* Instead the usual mode
of combining the reactants (furan and n-butyllithium) was reversed;
i.e* furan was added to the freshly prepared organometallic solution.
The cOTipletion of the metallation was carried out according to Benkeser and Currie*® to give a good yield of 2-furyllithium ( 7$-80^ ) as determined by Gilman's method of analysis**. The excess lithium metal then was removed from the reaction flask with a flat scoop.
22 H. Gilman and R.L. Bebb, J. Am. Chem.Soc., 109 (1939). 23 R.A. Benkeser and H. Landesman, J. Am. Chan. Soc., 7 ^ 2li93 (19U9). 24 H. Gilman and G.F. Wright, J. Am. Chem. Soc., %1, i S ^ (19U9). 2C R.A. Benkeser and R.B. Currie, J. Am. Chem, Soc., jg, I78O (19U8). -27-
Cyclopentanone (or cyclohexanone) was added slowly to the well-
cooled solution and the mixture was refluxed for several hours,
hydrolyzed, and worked up for product in the usual manner.
The distillation of each reaction product under reduced pres
sure (2 mm.) resulted in a considerable amount of dehydration* How
ever, the dehydration was accompanied by a negligible amount of
polymerization, as evidenced by the presence of very little residue.
The distillate from each of the above fractionations had to be
treated differently. In one case, the alcohol |^-( 2-furyl )cyclo-
pentanoiQ had separated cleanly from the olefin during the distil
lation. Therefore, the pure alcohol was dehydrated according to the
method of Bachmann and Heisey^^ and the olefin produced was combined
with that obtained during the distillation. In the second case the
alcohol j^-(2-furyl )cyclohexand^ had not separated from the olefin
during most of the distillation. Therefore, the entire distillate
was subjected to the dehydration method.
fields of the purified olefins (assumed to be l-(2-furyl)-
cyclopentene and l-( 2-furyl )cyclohexene ) were and $0% re
spectively (based upon 2-furyllithium). Since the distillation curve
for each dehydration product showed but a single flat and only one refractive index was obtained for all fractions collected, a single
2-furylolefin was assumed to have been produced in each case. The reduction of these compounds to l-( 2-furyl )cyclopentane and l-(2TftoiylV cyclohexane by sodium and liquid ammonia was accomplished in 95% and
97% yields, respectively. “28-
All new compounds encountered in the preparations outlined in
this section have been characterized by their physical properties
and elemental analyses (carbon and hydrogen). Adducts of final
products and maleic anhydride were prepared and these in turn were
characterized by elemental analyses (carbon and hydrogen) and melting
points *
The Production of Miscellaneous 2-Substltuted Furans
Furf uryliden e Diacetate and Fur furyl Acetate.— These two com
pounds were prepared from furfural and furfuryl alcohol according to
directions outlined in references 26 and 27, respectively, in yields
of '10% and 80^.
2-Bromofuran«— The procedure decided upon for the preparation of
this halofuran was essentially that of Shepard, Vfinslow and Johnson^®,
the first step of which was later supposedly improved by Moldenhauer and co-workers^s,
^-Bromo-2-furoic acid, prepared by brorainating 2-furoic acid in acetic acid, was obtained in 70^ yield by the present author after several discrepancies in the Moldenhauer's method®® were corrected;
1) three times the quantity of acetic acid prescribed was required in ae R.T. Bertz, Organic Syntheses. Vol. 33, John Wiley and 3 order to dissolve the 2-furoic acid, 2) the amount of bromine called for was in large excess and tended to lower the yield of product con siderably, 3) the rate of the reaction was considerably less than represented. The reaction was found to be mild and required no cool ing; a reflux period was necessary in order to insure complete re action and an optimum yield of product. The purified acid was decarboxylated by copper-bronze in quino line in 50-6OÎÉ yields to give 2-brcznofuran, Tetrametbylfuran.— This furan was synthesized by the two-step procedure of Gaertner and Tonkyn^®. 3,lt-Dimethyl-2,^"hexanedlone, prepared in 2^% yield by dlmerizing methyl ethyl ketone with di-tert- butyl peroxide^®, was cyclodehydrated by refluxing a mixture of it and acetic anhydride in the presence of a catalytic amount of an hydrous zinc chloride. Because of considerable polymerization, a yield of only li$% was realized in this second step. Production of Diels-Alder Adducts of 2-Alkyl- and 2-Methyl-^-alkyl- fur ans with Maleic Anhydride Stereochemical Considerations.— Ihe l,li-conjugate system of the nucleus in furan and its homologs is sufficiently dienic in character to permit ready participation in the Diels-Alder reaction with maleic 29 R. Gaertner and R.G. Tonkyn, J. Am. Chem. Soc., S072 (1951). 3® C.G. Moore, J. Chem. Soc., 1951. 236. -30- anhydride and several other dienophilic substances. A review of much of the published work with furan compounds is to be found in Norton's excellent summary of the Diels-Alder diene synthesis^^ and Dunlop and Peters^ book. The Furans^^. In the early work of Diels, Alder and co-workers^^'3 4 , it was assumed that the acid. A, obtained by the condensation of furan with maleic acid in aqueous medium was identical to the acid B obtained upon hydrolysis of the furan-maleic anhydride adduct prepared in ether. However, they were unable to isolate A from its solution due to facile dissociation into its generators vdiile, on the other hand, acid B could be purified by recrystallization from warm water. The above authors unfortunately arrived at their conclusions after carrying out a series of misleading experiments designed to determine the stereospecificity of reactions of furans with maleic derivatives. Upon treating the aqueous solution of A with bromine, a bromolactone was obtained. 'When B was dissolved in water and treated in the same manner it also yielded a bromolactone which showed the same empirical formula. Diels and Alder concluded that only one stereochemical form of acid could be present and assigned to it the endo configuration on the basis of the following interpretation of both reactions; 31 J.A. Norton, Chem. Revs., ^ 319 (19U2). 32 A.P. Dunlop and F.K* Peters, The Furans. Reinhold Publishing Corp., New York, N.Y., 1953. 33 0. Diels, K. Alder and E. Naujoks, Ber., 551; (1929). 34 0. Diels, et al., Ann., 1:90. 2h3 (1931)* 35 K. Alder and K. Backendorf, Ann., $3$. 101 (1938). -31- o O Hr Br CO,// HCt',, ro..ff -OH (D CO. I/) ■'.O. // Ô- In 191*8, Woodward and Baer^® demonstrated that in reactions of furan with maleic derivatives both stereochemical adducts could be formed, depending upon reaction conditions. Thus, furan and aqueous maleic acid give a product having the endo configuration whereas furan and maleic anhydride in ether react with the formation of the alternate exo adduct. In a classic piece of experimental work, the above authors proved that the acid B, when treated with bromine in aqueous solution, behaves in the following manner: cTO,// ^ H O B o Thus, the German workers were correct in their assignment of an endo configuration to A but B has an exo configuration. Of an un fortunate nature is the fact that they also assigned the endo con figuration to acids obtained upon hydrolysis of adducts prepared in ether from 2-methylfUran and 2, ^-dimethylfuran with maleic anhydride, based upon the same reaction cited above. 3* E. Tfocdward and H. Baer, J. Am. Chem. Soc., JO, 1161 (191*8). -32- It can, however, be logically assumed that furan homologs also react with maleic anhydride in ether to give the exo form of sub stituted 3,6-epoxy-^-tetrahydrophth&lic anhydrides. Propération of Adducts. —Only slight variations in method have been reported in the literature for preparing Diels-Alder adducts of substituted furans and maleic anhydride* Generally, it has proven sufficient to dissolve the reactants and a trace of hydroqulnone or other polymerization inhibitor in ether at room temperature and shake the solution for a period of time varying from several hours to a week. In the cases of the 2-substituted and 2,^-dlsubstituted furans, bulky substituents often necessitated heating the reactants in a solvent such as benzene or toluene. Tfhen the ether-solvent method was applied to the furans utilized in this research program, the adducts were obtained in a variety of crystalline modifications; rhombohedral clusters, large prismatic needles and loose bundles of fibres of different lengths were collect ed, their forms dependent someidiat upon the size of the substituent( s ) on the furan ring. Since relatively large quantities of a number of these adducts were required and in a finely divided state, consid eration of the physical forms in which they were obtained under the usual conditions was very important, especially when dealing with the first modification mentioned above* These crystals were ex tremely hard and required time-consuming repeated pulverizing and -33- screening operations. The necessity for finely divided adducts will be clarified when aromatization studies are discussed in the next chapter. In order to expedite the formations of these compounds and obviate the difficulty discussed above, several modifications of the usual procedure were made. The reactants, inhibitor, and three to four times their volume of ether were placed in a stainless-steel beaker and then warmed at a constant with hand stirring, until solution took place and an exothermic reaction ensued; in Arom ten to twenty-five minutes the reaction was complete. The warm solution was allowed to cool until no further evidence of crystallization was observed; the adduct then was collected by filtration* In several cases no evidence of reaction was observed and a warming period of two or three hours was allowed to compensate for the probable slowness of the reaction; the additional warning usually was successful and the adduct could be collected after over night refrigeration. Te trame thylfuran behaved in a much different manner than the others; it reacted so vigorously with maleic anhydride that cold-water cooling was necessary to moderate the reaction. 2-Methyl-5-n-propylfuran required special attention which will be described below. Ether had to be rqalenished during the above re actions, but an excessive amount was not lost. After removal of the adduct the filtrate was concentrated and cooled in an ice-salt bath in order to obtain an additional amount of product. In instances where large crystals would otherwise be formed, a motor-driven Hershberg-type stirrer was placed in the container at the first sign of crystallization and very rapid stirring applied for fifteen minutes to one-half hour as the solution cooled* As a re sult the adducts were obtained as very fine granular material, suit able for use in the aromatization step. The 2-alkyl- and 2-methyl-^-alkylfurans employed in this re search may be divided into four different classes, according to their rate of reaction with maleic anhydride and physical condition of the product: 1) those which react very rapidly with maleic anhydride and require rapid stirring in order to insure finely-divided states of adduct, 2) those which react very rapidly with maleic anhydride, but yield adducts of sufficiently fine physical form so as not to require stirring, 3} those which react rather slowly with maleic anhydride and require short warming periods^ the adducts in this class are all fine needles, U) those which react very slowly^ only example, 2-methyl-^-n-propylfuran. 2-Methyl-^-n-propylfuran had to be treated with maleic anhydride in refluxing benzene solution for several hours^ fine needles of the adduct then crystallized out when the solution was cooled. The substituted furans are classified in Table IV with the exception of 2-1ert-butylfuran. "35- Table IV Classes of Subetltutod Furan-Malelc Anhydride Adducts______ Compound______Class 1 Class 2 Class 3 Class U 2-Methylfuran x - - - 2-Ethylfuran x - - - 2-n-Propylfuran - x - - 2-Isopropylfuran - x - - 2,5-Dlm0thylfuran X - - - 2’-Methyl-5-ethylfuran X - - - 2-Kethyl-^-n-propylfuran - - - X 2-Furylcyciôpropane - x - - 2-Fury Icyclopentane - % - - 2-Furylcyclohexane - % - - 2-Bromofuran - - x - Furfuryl acetate - - x - Fur fury lidene dl acetate - - x - Tetrame thylfuran x - - - -36“ 2“tert-Butylfuran resisted all efforts to bring about its re action with maleic anhydride and with maleimide. A number of reaction temperatures were employed. Including a fairly high temperature in pressure equipment. Solvents were varied with no effect. Apparently the bulk of a tertiary alkyl group in the 2-position of furan pre vents adduct formation; no example could be found in the literature of such a Diels-Alder condensation when a group was larger in bulk than isopropyl. With the single exception of 2-methyl-6-n-propyl-3,6-epoxy-^'*- tetrahydrophthalic anhydride (the product from 2-methyl-^-n-propyl- furan and maleic anhydride), the pure addends were quite stable and upon standing for months at room temperature became only sli^tly discolored. It was observed, however, that if a compound was not initially colorless its rate of decomposition was markedly accel erated. In solution the adducts varied somewhat in stability; as a rule decomposition was accelerated. In most cases, whei these solu tions were heated above 50-6o*, total reversion of the adducts into their generators took place. Very high yields of products were realized in all of the con densations. Since the adducts were quite pure only sufficient amounts necessary for characterization purposes were recrystallized. These recrystallizations were effected by dissolving the compounds in chloroform at room temperature, adding petroleum ether (60-80* ), then cooling the solution slowly in an ice chest. -37- All adducts were characterized by their melting points and, when new, by elemental analysis (carbon and hydrogen). In addition, char acterization of the phthalic anhydrides from the adducts which were capable of aromatization served as additional confinnation. The Production of 3-AlkyIphthalic Anhydrides and 3-Methy1-6-alkyl- phthalic Anhydrides General Background. — Very few examples exist in the chemical literature of transformations of furan- or substituted furan-maleic anhydride condensation products into aromatic systems through the agency of acid catalysts. The first of these was reported in the form of a note in 1933 by Van Campen and Johnson®"^. These investi gators heated glacial acetic acid solutions of 3-methyl-, 3-bromo- and h-bromo-3,6-epoxy-/^-tetrahydrophthallc anhydrides in the pres ence of hydrogen bromide and obtained the corresponding 3-methyl-, 3-bromo- and L-bromophthalic anhydrides. No conditions of their experimental work or yields of products were given. In 19UUJ Newman and Lord^^ prepared the 3,6-dimethylphthalic anhydride from 3,6-dime thy 1-3,6-epoxy-/::^-tetrahydrophthalic anhydride by slowly introducing the powdered Diels-Alder adduct, with stirring, into cold 90^ sulfuric acid. After pouring the dark-red solution over ice, the expected anhydride in yield and a small amount of 2,5 -dimethylben2oic acid were isolated. 37 M.J. Van Campen and J.R. Johnson, J. Am. Chem. Soc., ^ U30 (1933)* 3s M.S. Newman and B.T. Lord, J. Am. Chem, Soc., ^ 73U (19Uii). It was not until eleven years later than mother such arouiatisa- ticn was reported. Sherman and Dunlop^® obtained yield of 3“n:ethyl-6-Eiethcxyphthalic anhydride by hoatin;^ tlic corresponding adduct in glacial acetic acid in the presence of a trace of anhydrous zinc chloride. A year later Cava and c o - w o r k e r s , i n order to characterize 2-acetoxyfuran, prepared its maleic anhydride adduct and converted it into the known 3-acetoxyphthalic anhydride in $7% yield by briefly heating its acetic anhydride solution at 120° in the presence of a little sulfuric acid. Tlie undertaking of the present research program was encouraged ly the first two reports cited above, and was predicated upon the feasi bility of carrying out this step of the over-all scheme in acceptable yields with a number of addition compounds* The facile preparation of these adducts in exceptionally high yield and purity from their precursors, the ready availability aid low cost of maleic anhydride, and the ease with which many furans can be prepared (or obtained commercially), made the aromatization step appear the crucial one* An investigation was therefore carried out in order to determine which acid catalysts would best effect the aromatization of the adducts derived from 2-methylfuran and 2,5“dimethylfuran, respective ly, aid maleic anhydride. E. Sherman and A*P* Dunlop, lk7th Meeting of The American Chemical Society, Cincinnati, Ohio, 1955* 40 H.P. Cava, C.L. Wilson and C*J* Williams, Jr., J* Am. Chem. Soc., la, 2303 (1^56). -39- HydroRen Bromide Method.— The work of Van Campen and Johnson, as previously pointed out, is incorporated into the literature in the form of a note. The lack of experimental detail therefore necessi tated a thorough examination of reactant concentrations and reaction conditions j for tliis purpose the adduct of 2-methylfuran and maleic anhydride was chosen initially because of the success claimed by those investigators in its aromatization* Upon the assumption that anhydrous hydrogen bromide in a glacial acetic acid medium was employed by Van Campen and Johnson, a series of experiments was performed in which the quantity of catalyst was varied fixim a trace amount to a molar excess and reaction temperatures of 30® to 120“ were usedj the amount of adduct and solvent were maintained constant. A general darkening of the solution resulted in each case, either slowly of rapidly depending upon the temperature and concentration of catalyst. The end product always was the same— a brown polymeric material containing none of the desired anhydride. In a second group of eiç) eriments, concentrated hydrobromic acid (UB%) was introduced into addend solutions in quantities similar to the ones used above j the effect of temperature on this system also was studied. A rapid darkening of solution always resulted, and in from ten minutes to one day a light-brown finely*»divlded solid separated from the acetic acid solution^ the time of separation depended upon whether high or low temperatures were ^plied. This substance could -Lo be readily purified by vacuum sublimation or recrystallization from its hot aqueous solutionj however, it proved to be fumaric acid and not 3-methyIphthalic anliydride or 3-methylphthalic acid. The acetic acid filtrate yielded only brown polymer iriien it was worked up in the usual manner. Reactions similar to those described above also were carried out with 3,6-dime thy 1-3,6-epoxy-^-tetrahydrophthalic anhydride in acetic acid. The results were the same; anhydrous hydrogen bromide caused only polymerization while hydrobromic acid produced fumaric acid and polymer. The aromatization experiments described above are outlined in Table V. -Ill- Table V ——e-3-c==sss™ B2œ e!^a*===ââG*âGw êâ*=G eS==^BEi^=^=M : Temp. & Time Compound (Moles) Acid (Moles)______of Reaction Result 3-la'ethyl-3,6-epoxy- Anhyd.HBr( trace^ 120" for 1 hr. Polymer only A4-tetrahydro- Anhyd. KBr(O.Ol) 30" for 15 hrs. Polymer only phthaiic anhydride Anhyd. HBr(O.Ol) 120" for 1 hr. Polymer only (0.1) in 250 ml. Anhyd. HBr(O.l) 30" for 15 hrs. Polymer only glacial acetic acid Anhyd. HBr(O.l) 90" for 2 hrs. Polymer only Anhyd. HBr(0.2) 30" for 15 hrs. Polymer only Anhyd. HBr(O.S) 90" for 2 hrs. Polymer only Aq. HBr (trace) 30" for 2h hrs. Polymer only Aq. HBr (0.01) 30" for 2\x hrs. Polymer; brown powder, not purified Aq. HBr (O.Ol) 90" for 2 hrs. Polymer; a little fumaric acid Aq. HBr (0.02) 30" for 2U hrs. Polymer; a little fumaric acid Aq. HBr (0.05) 90" for 2 hrs. Polymer; fumaric acid Aq. HBr (O.l) 60 " for 5 hrs. Polymer; fumaric acid Aq. HBr (O.l) 90" for 2 hrs. Polymer; fumaric acid Aq. HBr (0.1) 120" for 1 hr. Polymer; fumaric acid Aq. HBr (0.5) 90" for 2 hrs. Polymer; fumaric acid 3,6-Dimethyl- Anhyd.HBr(trace) 120" for 1 hr. Polymer only 3, 6-epoxy-A‘*-tetra- Anhyd.HBr (O.l) 30" for 15 hrs. Polymer only hydrophthalic Anhyd. HBr (0.1) 90" for 2 hrs. Polymer only anhydride (0.1) in Anhyd. HBr (0.1) 120" for 1 hr. Polymer only 250 ml. glacial Aq. HBr (0.1) 30" for 2U hrs. Polymer; acetic acid fumaidc acid Aq. HBr (O.l) 90" for 2 hrs. Polymer; fumaric acid Aq. HBr (0.1) 120" for 1 hr. Polymer; fumaric acid Aq. HBr (0.5) 90" for li hrs. Polymer; fumaric acid -L2- SulfttJic A d d Method»— The second method 'iriiich reportedly furnished a phthalic anhydride when the corresponding furan-maleic anhydride adduct was treated with an acid is that of Newman and Lord^°. When their procedure was repeated with 1,6-dlmethyl- 3.6-apo3gr-/^-tetrahydrophthalic anhydride, a creditable yield of 3.6-dimethyIphthalic anhydride was obtained. In order to ascertain the factors determining optimum yields of anhydrides, the above adduct and 3-aethyl-3,6»epoxy-^-tetrahydro- phthalic anhydride were aromatized in sulfuric acid of concentrations ranging from $0^ to 96$ by weight and at temperatures of -1^® to +10“• The effect of acid-adduct ratios upon yields of anhydrides was examined, also. A brief outline of the general procedure which was followed is given below; To a suitable quantity of sulfuric acid, cooled to the desired reaction temperature, was added powdered or finely granular adduct, with rapid stirring, at such a rate as to allow complete wetting of each portion as added. TSiis rapid stirring and the cooling were maintained after addition was completed until a clear orange-to-red solution was attained. When the solution reached a predetermined temperature it was poured over cracked ice; the solid material which separated was filtered and washed with cold water (the filtrate, upon being cooled with an ice-salt mixture, yielded no appreciable -ii3- amount of additional material). The wet residue then was stirred in an ice-cold sodium bicarbonate solution for a short time in order to free the anhydride of any methyl- or dimethyIbenzoic acid or colored impurities, and then was refiltered. A light-tan to white product resulted. An examination of the filtrate from the bicarbonate washing was made for a number of runs of each adduct. Upon acidification no 3,6-dimethylphthalic acid was ever obtained; in a few cases to be noted later, 3-methylphthalic acid was obtained in small quantities when the corresponding adduct was aromatized in quite dilute sulfuric acid solutions. Of special interest was the observation that the only benzoic acid formed during the aromatization of this adduct was m-toluic acid. Reactions of other monoalkylfuran-maleic anhydride adducts were later found also to yield only the meta-alkyIbenzoic acids. The use of powdered or finely divided adduct was found to be absolutely necessary for smooth reaction; otherwise the adduct did not dissolve cosqjletely in the acid, and when the solution was hydrolyzed the product contained reddish-colored bits of unreacted adduct (which rapidly resinified and could not be removed easily). Usually the anhydrides obtained in this manner were reasonably pure and could be used directly for the next step of the reaction sequence. For the purpose of obtaining accurate yield data, however. the two anhydrides discussed above were recrystallized from a hot benzene-Skellysolve B mixture, after decolorization by activated charcoal. The results of these experiments are summarized in the following table. -h$- Table VI Sulfuric Acid Aroinatization Fjcoeriments Amount of Reaction and Hydrolysis Yield of Cotnpound( .Amount ) Acid & Cone. ______Temperatures______Anliydride 3-I.!ethyl-3,6- epoxy-^~tetra- hydrophthalic anhydride (20 g.) 2^0 ml., S 0 % No dissolution until 10® 5.0%, in immediately poured over cluding ice some diacid (20 g.) 2^0 ml., ^ 0 % Dissolution at 10®, 6.0%, in stirred overnight at cluding room temperature, much diacid poured over ice (30 g.} 350 ml., 75% Dissolution at -10® to 20.0%, and -5®; allowed to stand 6 5*0%. diacid hrs. at 5"5 heavy mass of crystals settled out; poured over ice at 5® (30 g.) 350 ml., 60% Dissolution at -10® to 33.0-36.0%, -5*3 allowed to stand and 5.0- 13 hrs. at -5*3 heavy 7.0% diacid mass of crystals settled out; TDOured over ice at -5* (30 g.) 350 ml., 85% Dissolution at 10®, 35.0-37.0%, allowed to stand 2h hrs. and 5*0—6.0% at 0*3 heavy mass of diacid crystals settled out; poured over ice at 0® (30 g.) 350 ml., 90% Dissolution at -15* to ii2.5-U5.5% -10*3 stirred for 1 hr. longer and allowed to rise to 0®, then poured over ice (30 g.) 350 ml., 90% Dissolution at -10® to li2 .0—Jtii .0% -5*3 allowed to rise to 0® j then poured over ice (30 g.) 350 ml., 90% Dissolution at -5* to 33.0% 0®, immediately poured over ice; some SOj fumes detected -U6- Table VI (Continued) Amount of Reaction and Hydrolyuis Yield of Compound(Amount ) Acid & Gone.______Temperatures Anhydride (30 n*) 200 ml., 9 0 % Dissolution at -15" to h h , $ % -10°} stirred for 1 hr. longer and allowed to rise to 0 ", poured over ice (60 g.) IdO ml., 9 0 % Dissolution at -15" to *0% -10“j stirred for 1 hr. longer and allowed to rise to 0 “, poured over ice (30 g. ) 350 ml., 9 à % Dissolution at -10“ to h 9 » 0 - h 6 , O % -5 "j allowed to rise to 0 ", poured over ice (30 g.) 350 ml., 96% Dissolution at -5“ to O" 37.0% immediately poured over icej some SOg detected (60 g. ) 180 ml., 96% Dissolution at -15" to Uii.5% -10“j immediately poured over ice 3.6-Dimethy1- 3.6-epoxy-^4- tetrahydrophthalic anhydride (30 g. ) 350 ml., 80% Dissolution at 0°} 2h 32.0% hrs. at room tempera ture j mass of crystals settled; poured over ice at 30“ (30 g. ) 350 ml., 90% Dissolution at -15" to h7*5-50.0% -10"; allowed to rise to 5", poured over ice (30 g. ) 350 ml., 90% Dissolution at -10“ to h6.0-1x8.0% -5"; allowed to rise to 5", poured over ice (30 g.) 350 ml., 90% Dissolution at -5“ to h6.0-1x9.0% 0"; allowed to rise to 5", poured over ice (30 g.) 350 ml., 96% Dissolution at -15" to 50.0-53.5% -10"; allowed to rise to 5", poured over ice -1(7- Table VI (Continu« ) Amount of îienction and Hydrolysis Yield of Compound(Amount) Acid & Cone, Temperatures Aniivdride (30 g.) 350 ml., 96^ Dissolution at -10° to h 9 ,o - 9 i,5 % -5°; allowed to rise to 5 ", poured over ice (30 g.) 350 ml., 96% Dissolution at -5“ to 51.0-52.0% 0°; allowed to rise to 5 °, poured over ice (àO g. ) 350 ml., 96% Dissolution at -15° to 50.5% -10°, stirred for 1 hr. longer; allowed to rise to 5 °, poured over ice (100 g . ) 350 ml., 96% Dissolution at -15° to 52.0% -10°; stirred for 1 hr. longer; allowed to rise to poured over ice In addition to the above expeitLments, several large runs were made using a kilogram or more of each adduct. Because sulfuric acid (commercial or C.P.) at -1^° to -10“ gave optimum yields of products in the smaller runs, the same acid concentration and temper ature range was employed here also. Essentially the same procedure was followed, with the resultant formation of 3 -methylphthalic and 3 ,6-dime thy Iphthalic anhydrides in kO-h$% and $(>•$$% yields, re spectively. The results of the application of the modified Newman and Lord procedure to other Diels-Alder adducts (substituted furans and maleic anhydride)are summarized in the table below. Yields are of purified nroducts. -tid- Table VII I'arious Sub: rvran-iîaleic Anhydride Amount of Hydrolysis Yield of Adduc t (Amount ) Acid Temperature Anhydride 2-Ethyll'uran (^0 g.) 375 ml. -lx = 1x3.6% (SO g.; 375 ml. 0 ° 56 .0% (50 g.) 375 ml. 3° 1x7.0% (50 g.) 375 ml. 5" h9.li% (375 g.) 2.3 1. 5° 1x7.5% (725 g.) 3.8 1 .-2 " 53.0% 2 - n-Propy If ur an (50 g.) 375 ml. 5" 32.5-Uo .o % (333 g.) 2.33 1. 10“ 35.1i% (330 g.) 2.33 1. 0 “ h k . 2 % (350 g , ) 2 .1x 1 . 0 “ 1x1.5% 2-Isopropylfuran (1x0 g. ) 375 ml. 5° 1 0 . 0 % (100 g.) 500 ml. 0 “ 9.5% f:*: (200 g.) 1 1 ., 9 0 ^ 0 “ 8.5% ^ 2-Methyl-^- ethylfur an (50 g.) 375 ml. 10" 61.0%, 67.0% (Ii50 g.) 2.5 1 . 2" 57 .7% (1x62 g. ) 2.5 1 . 5" 60.5% 2-Me thyl-^-n-propylfuran (50 g.) 375 ml. -7° 1x7.0% (50 g.) 375 ml. 6" 1x6.0%, 1x8,0% (iiO g.) 375 ml. 0" 1x1.2% (Lo g.) 375 ml. 5" 38.7% (700 g.) Ix.O 1. 5” 1x0.5% Tetramethylfuran (50 g.) 375 ml. 5“ 5 .0% (50 g.) 375 ml. 10" 5 .0% 2-FuryIcyclopropane (30 g.) 200 ml. 0" 0 .0%, dark polymer (30 g.) 200 ml,90% -5° 0.0%, dark polymer '^Exceptions are in colnmn after amount of acid. •^•J^'otained as 3-isopropylphthalic acid. -hP- Table VII (Continued) Furan-Maleic Anhydride Air.ount of Hydrolysis field of Adduct (Amount) Acid Temperature Anhydride 2-Fury Icyclopentane (go g. ) 375 ml. -10°, froihing O.C^, yellow was considerable polymer (50 g.) 375 ml. 90^, -10° O.C^, yellow acid polymer (50 g.) 375 ml. ao$, -5“ O.O^j much acid of adduct recovered 2-FuryIcyclohexan e (50 g. ) 375 ml. 5° 0.0^, pink polymer (50 g.) 375 ml. 90^ 0“ 0.0^, pink polymer 2-Eromofuran (50 g.) 375 ml. 5*, HBr evolved 0.0^, dark rapjdly during polymer aromatization (50 g.) 375 ml. 90^ 0°, HBr evolved O.G^, dark during aromati polymer zation 2-Fur fury lidene diacetate (50 g.) 375 ml. -5“> odor of 0 , 0%3 dark furfural polymer (50 g.) 375 ml. 80^ -5°, odor of 0,0%3 dark furfural polymer 2-Furfuryl acetate (50 g.) 375 ml. -5“ O.C^, dark polymer (50 g.) 375 ml. 90^ -5" 0.0%, dark polymer In all cases where successful aromatizations were realized, additional amounts of the phthalic anïiydrides were prepared (but not purified) for reductions to the hylrocarbons (exceptions: the yields of 3-methylphthalic acid and tetramethyIphthalic anhydride were toe low to justify further consideration). -50- The possible mechanisr. of these aromatrisations may be illus trated, using the 2 -aethylfuran-maleic anhydride adduct (exo fora} as example, in the following manner: c - c C o HO H C-0 J o CH CH C -O c-o HO C-O HjO -o C-o In addition to the anhydride, the 2,5“dimethylfurar.-, 2-methyl- furan-, 2-ethylfuran- and 2-n-propylfuran-maleic anhydride adducts also gave rise to isolable m-alkylbenzoic acids in small quantity Tihen aromatized* It was proven conclusively that the decarboxylation took place during the aromatization because when the pure anhydrides were re-dissolved in S 6 % sulfuric acid, even at room temperature, and then the solutions were poured over ice, no alkylbenzoic acids were found. This loss of carbon dioxide can logically take place in the following manner, starting from the third formula above: CH, H© 1^0 hydrof. CO. // “^1“ Probably a combination of electrical and steric factors accounts for this behavior, yihen the epoxy ring opens, the carbonium ion will most likely be formed with the positive charge at the carbon atom to which the aIky1-group ia attached. The formation of a double bond then can take place if either a proton or carbon dioxide is elimi nated (see previous page). In addition to the two methods described, numerous other expei*i- ments were performed in which acids of various types were used in attempts to bring about aromatization of the two adducts designated as test compounds in the previous work recorded. Some of these acids served also as the solvent^ others were dissolved in appropriate sol vents in various quantities ranging from trace amounts to large ex cesses. A few amine salts also were tried as aromatization agents. In order to simplify recording of the results of these experiments, they have been placed in Table Till. -52- Table VIII Other Aromatir.ation E:cpcriments Compound Catalyst Temp. & Time (I'olns } (Amount) Solvent of Reaction Result 3-Hethyl- 6 - e p o x y - ^ - tetrahydro" phthalic an hydride (0.1 j Anhyd. HBr Ace' ic an 86-90" for Polymer (trace) hydride 2 hrs. only (O.lj Anhyd. HBr Acetic an 60“ for 2 Polymer (0.1 mole) hydride hrs. only (0.1) Anhyd. HCl Glacial 60“ for 5 Polymer (approx. 0.1) acetic acid hrs. only (0.1) Aq. cone. HCl - 25-30“ for Polymer (100 ml.) 5 hrs. only (0.1) Acetyl chlor Glacial 25-30“ for Polymer ide (2 ml.) acetic acid 12 hrs. only (0.1) Acetyl chlor Glacial 80-90“ for Polymer ide (2 ml.) acetic acid 2 hrs. only (0.1) Polyphosphoric - 90-95“ for Polymer acid (200 ml,) 1 hr. only (0.1) Phosphoric 85-90“ for Polymer acid (85^) 1 hr. only (200 ml.) (0.1) Trifluoro- Glacial 90“ for 1 Polymer acetic acid acetic acid hr. only (2 ml.) (0.05) Formic acid — 88-90“ for Polymer (98-1005^) 1 hr. only (50 ml.) (0.1) D-Toluene Glacial 80-90“ for Polymer sulfonic acid acetic acid 2 hrs. only (1 g») ~53" Table VIII (Continued) Compound Catalyst Temp. & Time (Holes) (Amount) Solvent of Heaction Result 10.1 ) ^Toluene Acetic an 85-90" for Polymer only sulfonic acid hydride ? hrs. (1 g.) (0 .1 ) 9 6 % H2SO4 Acetic an 25-30" for Polymer and (0.25 ml.) hydride 12 hrs. some 5-œethyl- 2-fur3'l methyl ketone (0.1) 96% H2EO4 Acetic an 90" for 1 Polymer and (0.25 ml.) hydride hr. a little ketone (0 .1 ) Dichloroacetic Glacial 90“ for 1 Polymer only acid (2 g. ) acetic acid hr. (0 .1 ) Oxalic acid Water 90-95' for Polymer only (5 g.) 1 hr. (0 .1 ) Piperidine. Water 90-95" for Slow dark HCl (5 g.) 1 hi'. ening, polymer only (0 .1 ) Dimethylanine• Water 90-95" for Slow dark HCl (10 g.) 1 hr. ening, polymer only (0 .1 ) Anhyd. ZnClg Acetic an 25-30" for Polyner and (trace ) hydride 2h hrs. some ketone (0 .1 ) Anhyd. ZnClg Acetic an 25-30" for Polymer and (0.1 moles) hydride 2h hrs. some ketone (0.1) Anhyd. ZnClg Acetic an 120" for Polymer only (trac e) hydride 1 hr. ^This compound was identified through its semicarbasone in this experiment only. In all others only the detection of an oil -with its odor was recorded. -5h~ Table VIII (Continued) Compound Cotalys t Temp. 6 Time (bf'ies) _ (Amount} Solvent of Reaction Result (n.l) Arüiyd. InClj Claclal 90“ for 1 Polymer, (trace ) acetic ecid-■ hr. trace of acetic an ketone hydride (0.1) Anhyd. ZnClg llacial 120" for Polymer only (trace) acetic acid 1 hr. (0.1) Anhyd. FeClj Acetic an 120° for Polymer only (trace) hydride 1 hr. (0.1) Anhyd. FeCl^ Acetic an 25-30 ° for Polymer and (0.1 g.) hydride 2li hrs. some ketone (0.1) Anhyd. FeCl^ Glacial 90-95* for Polymer, (0.1 g.) acetic acid 1 hr. little ketone (0.1) Anhyd. SnCl^ Acetic an 25-30 ° for Polymer and (trace) hydride 2ii hrs. some ketone (0.1) Anhyd. SnCl^ Glacial 90° for Polymer only (trace) acetic acid 1 hr. (0.1) BF^-etherate Glacial 9 0° for Polymer only (2.0 g.) acetic acid 1 hr. (0.1) Anhyd. AICI3 Acetic an 90° for Polymer, (trace) hydride 1 hr. trace of ketone (0 .1 ) Anhyd. AICI3 Glacial 120° for Polymer only (trac e) acetic acid 1 hr. (0.1 ) Anhyd. SbClg Glacial 9 0° for Polymer only (0 .1 g . ) acetic acid 1 hr. (0 .1 ) P^Os (10 g.) Benzene 90° for Polymer only 1 hr. ”55 •• Table VIII (Continued) Compound Catalyst Temp. & Time (Moles) (Amount) Solvent of Reaction Result (0.1) BF^-etherate Glacial 120“ for Polymer only (1.0 g.) acetic acid 1 hr. 6-dimethy1- 3,6-epoxy-^4- te braliydro- phthalic an hydride (C.l) Anhyd. HBr Acetic an 25-30“ for Polymer only (trace) hydride 2k hrs. (0.1) Anhyd. HBr Acetic an 90“ for Polymer only (O.l mole) hydride 1 hr. (0.1) Anhyd. HCl Glacial 90“ for Polymer only (0.1 mole) acetic acid 1 hr. (0.1) Aq. conc. * 60“ for Polymer only HCl L hrs. (0 .1 ) Conc. Aq. Glacial 90-95“ for Polymer only HI (0.1 m o l e ) Acetic acid 1 hr. (0 .1 ) ^-Toluene- Glacial 90-92“ for Polymer only sulfonic acid acetic acid 1 hr. (2 g.) (0.1) P2O5 (10 g.) Benzene 90-95“ for Polymer only 1 hr. (0.1) Dowex 50 Glacial 90-95“ for Polymer only resin (2 g.) acetic acid 1 hr. (0.1) Trifluoro- Glacial 90“ for Polymer only acetic acid acetic acid 2 hrs. (2 g. ) (0.1) Acetyl Glacial 25-30“ for Polymer only chloride acetic acid 18 hrs. (2 ml.) -56- Table VIII (Continued) Compound Catalyst Temp. Ù. Time (Holes) (Amount ) Solvent of Reaction Result (0.1) Acetyl Glacial 60° for Pol'mer onlv chloride (2 ml.) acetic acid 2 hrs. (0.1 ) Polyphosphoric 85-90° for Polymer only acid (200 ml.) 2 hro. (0.1) Phosphoric 65-90“ for Polymer only acid (85;g) 2 hrs. (200 ml.) (0.1) Formic acid 60“ for Polymer only (98-100%) 5 hrs. (200 ml.) (0.1) 96% HgSO. Acetic an 90“ for Polymer only (2 drops) hydride 1 hr. (0.1) 9 6 % H2SO4 Acetic an 60“ for Polymer and (0.25 ml.) hydride 2 hrs. oil'^ (0.1) Piperidine Vlater 90-9^' for Polymer only .HCl (5 g.) 1 hr. (0.1) Dimethylamine Water 90-95" for Polymer only .HCl (10 g.) 1 hr. (0.1) Anhyd. ZnClg Acetic an 25-30“ for Polymer and (trace) hydride 18 hrs. o i l ^ (0.1) Anhyd. ZnClg Acetic an 90-95" for Polymer, (trace) hydride 2 hrs. trace of oil (0.1) Anhyd. ZnClg Glacial 120“ for Polymer onl}' (trace) acetic acid- 1 hr. acetic an hydride Yias not identified— probably was 2,5-dimethyl-3-furyl methyl ketone . -57- Table VIII (Continued) Compound Catalyst Temp. & Time (Moles) (Amount) Solvent of Reaction Result (0.1) Anhyd. ZnCl^ Glacial 90' for 1 Polymer only (o.l mole) acetic acid hr. (0.1) Anhyd. FeCl^ Acetic an 25-30' for Polymer, (trace) hydride 15 hrs. small amt. oil^ (0.1) Anhyd. FeCl^ Acetic an 90* for Polymer, (trace) hydride 2 hrs. small amt. oll^ (0.1) Anhyd. FeClg Glacial 120' for Polymer only (0.1 g.) acetic acid 1 hr. (0.1) Anhyd. SnCl* Acetic an 25-30* for Polymer, some (trace) hydride 2ii hr a. oil ^ (0.1) Anhyd. SnCl* Glacial 120' for Polymer only (trace) acetic acid 1 hr. (0.1) Anhyd. AICI3 Glacial 90' for Polymer only (trace) acetic acid 1 hr. (0.1) Anhyd. SbClg Glacial 90* for Polymer only (trace) acetic acid 1 hr. (0.1) BFg-etherate Acetic an 90 " for Polymer, oll'^ (2.0 g.) hydride 1 hr. (0.1) BF^-etherate Glacial 120* for Polymer only (1.0 g.) acetic acid 1 hr. ^ Was not Identified— probably is 2,$-dimethyl-3-furyl methyl ketone • All anhydrides (and 3-l5opropylphthalic acid) were characterised by their melting points and neutral equivalents. Elemental analyses (carbon and hydrogen) of all new anhydrides (and their methyl and/"or ethyl and/or butyl esters) were obtained. Physical properties of the esters (boiling ranges and refractive Indices) also were recorded. Reduction of 3-Alkyl- and 3-Methy 1-6-alkyIphthalic Anhydrides to Hydrocarbons General Background «— An intense desire to synthesize prehnitene (l>2,3jii-tetra!”ethylbenzene ) and hemimelliteno ( 1,2, 3-trimethyl- benzene) by the route proposed, and also the presence of plentiful quantities of 3f6-dimethyIphthalic anhydride and 3-methyIphthalic anhydride from the aromatization investigations resulted in the application of all early reduction experiments toward the acquire ment of those hydrocarbons* An obvious and convenient method of converting the anhydrides to the hydrocarbonswas considered first; the catalytic hydrogenation— hydrogenolysis of the dicarbalkoxy moitiés of the derived dialkyl phthalatesto methyl groups in the presence of a copper chromite catalyst. Such a reduction had been realized by Wojcik and his co- workers^^ in 1933 Tfhen they reduced diethyl phthalate to o-xylene in 8!;% yield, using a copper chromite catalyst stabilized against re duction by barium oxide. The reaction was slow, for 2-3 hours were required to reduce 0.1 mole of diester at a temperature of 250" and a hydrogen pressure of 100-170 atmospheres. However, in the present research both diethyl and dibutyl esters of the anhydrides selected for experimentation were prepared and reduction experiments were carried out. A second possible method of converting an anhydride to a hydro carbon was examined: reduction of the anlQrdride or diester to the 41 B. Wojclk, L.TT. Covert and H. Adkins, J. Am. Chem. Soc., 1669 (1933). -58- -59- dicl by lithium aluminum l^dride, preparation of the dibromide and finally, reduction of the dibromide to the hydrocarbon by a lithium hydride-lithium aluminum hydride mixture. W.G. Brown and his co- workers*®, in their early studies of the versatility of liüilum aluminum hydride, had reduced phthalic anhydride to o-xylylenol in 8756 yield. Benzyl bromides had been prepared in high yields frcan the alcohols through the use of phosphorus tribromide, and the re duction of benzyl halides by the method proposed had been employed by Shacklett and Smith*^ successfully in the preparation of a number of hydrocarbons. Mille the evaluation of this procedure was in progress a potentially valuable modification of it was conceived; if the diol could be reduced directly to the hydrocarbon by sodium and methanol in liquid ammonia, the dibroraide preparation would be elimi nated from the procedure. Birch's method'*^ of reducing primary benzyl alcohols directly to methylbenzenes ( in good yields ) by sodium and methanol in liquid ammonia was well known; however, no reference could be found in the literature irtdLch described the reduction of dlols of the benzyl type directly t>o the hydrocarbons* A thorough investiga tion of all possibilities was made. Die final method considered was predicated upon reports that 1) ethyl phthalate exhibited no reaction with liquid ammonia*® and *2 wJg. Broro, ^ al., J. Am. Chem. Soc., ^ 1197 (19U7). « C.D. Shacklett and H.A. Smith, J. Am. Chem. Soc., 23, 766 (1951). *4 A.J. Birch, J. Chen. Soc., 15U5 . 809. *® E. Bartow and D.F. McFarland, Rev. Am. Chen. Research, 8, 303 (1902). -60- 2 ) that sodium and alcohol had been used to reduce certain aliphatic esters to alcohols in liquid ammonia”*®. If this reduction could be applied to phthalates, and if the quantity of sodium and alcohol could be Increased sufficiently to continue reduction past the alcohol stage, a vastly superior way of arriving at hydrocarbons would be at hand. Several esters of this type were subjected to the reductive conditions described. The results of studies of the above-outlined means of access are described in the following sections. Copuer Chromite Method.— Adkins and Folkers"*"^, in their many in vestigations of ester reduction, concluded that butyl esters are more easily reduced than ethyl esters, whereas methyl esters are quite difficult to reduce. Accordingly, both dibutyl and diethyl 3-methyl- phthalate and 3,6-dimethylphthalate were prepared; these esters were given careful vacuum distillations and then redistilled from a small amount of Raney nickel in order to remove any sulfur-containing im purities which might have been carried over from the estérifications of the anhydrides. Two copper chromite catalysts were employed; these were fresh batches, purchased from the Harshaw Nickel Company, having the following compositions; (a) 39jS cupric oxide, kl% chromic oxide, 22$ barium oxide; (b) h2% cupric oxide, chromic oxide, and 1C$ 46 E. Chablay, Compt. rend., 1S6. 1020 (1913). 47 H. Adkins and K. Folkers, J, Am. Chem. Soc., ^ llU5 (1932), -61- barium oxide. All reductions were carried out in a 300-ml. stainless steel bomb, seated in a rocker carriage which contained its own heating element^ temperature ciontrol was easily accomplished by means of a thermocouple and an electrical regulator. In order to attain the high hydrogen pressures necessary, a booster pump was used in con junction with the usual system linking the hydrogen tank to the autoclave. The bomb was normally charged with 0 ,$ mole of dieater and 10-20^ by weight of catalyst (with or without solvent) and pressured to between 100 and 1^0 atmospheres; by the time a tempera ture of 2$0-300* was attained, reaction pressures of 250-li00 at mospheres were reached also and then were maintained by means of the booster pump. Upon termination of the reaction, the catalyst was removed by filtration and the mixture was vacuum distilled. Unfortunately, only traces of prehnitene were obtained when dleijiyl or dibutyl 3,6-dimethylphthalate were reduced. Variation of catalyst amounts, solvents, and reaction ccmditlons ( including step wise reduction) always resulted in a pred(minant amount of cleavage to 3,6-dlmethylphthalic anhydride and the formation of lesser amounts of ethyl (or butyl) 2-hydroxymethyl-3,6-dlmethylbenzoate and 3,li-benao- 2,5 ■'dihydro fur an. Other products present could not be separated by vacuum distillation. It is interesting to note that no diol was isolated. It is a solid and distils at a high tei^perature under vacuum and undoubtedly —62— Trould have been present in the residue of the distillations; the presence of 3#U“benzo-2,^-dihydrofuran could very well indicate that the diol stage was reached but that dehydration took place under the conditions of the experiment. At any rate, stepwise reduction was not successful, for no hydrogen uptake was detected until temperatures of over 225® were reached, and thereafter it was extremely slow* The formation of hemlaellitene by the reduction of the j-nethyl- phthalates was accomplished by hydrogenolysis over copper chrcsaite, but the method left much to be desired. In order to convert 0,5 mole of ester into hydrocarbon, periods of time as long as 36 hours were required; an optimum yield of 78% was obtained when the dibu-tyl ester and 30% by weight of catalyst were used, without solvent, and a temperature of 275* and a pressure of 300 atmospheres were maintained for 2h hours* The copper chromite method therefore was concluded to be of little value in the present research. It should be noted that no difference in the activity of the two catalysts was detected. Reduction of Anhydrides to Hydrocarbons. — %hen 3,6-dimethyl- phthalic anhydride was reduced by lithium aluminum hydride, the diol was not obtained* The infra-red spectrum of the product Indicated that it was 3,6-dimethylpbthalide. Elenmntal analysis (carbon and i^drogen) of the compound supported this assignment of structure. The phthslide was obtained in 8d-90% yields. Upon subjecting the —63” lactone to the same reduction procedure, 3>6-diiaethyl-o-xylylenol was produced in 9056 yields. Apparently the methyl groups of the anhydride sterically inhibited the formation of the bulky reduction intermediate which would have resulted if both carbonyl functions bad been reduced simultaneously. Hydrolysis of the initial reduction mixture and a repeated reduction of the product thus obviated this difficulty. The preparation of the dibromide from the diol was effected in 95% yield, and the final reduction was carried out in 7^% yield to give prehnitene. Some time after this synthesis had been achieved, Buchta and Loow^® reported it in exact detail. However, this method leaves much to be desired since all three reductions re quired the use of a Soxhlet outfit and only small quantities of materials could be reduced at once. 3-Methylphthalic anhydride was found to undergo direct reduction to 3-niethyl-o-3cylylenol in 8)56 yield, but only one small run was carried out. The diol was not converted into the bromide, because at this point in the investigation a method of converting it directly to the hydrocarbon was found. Reduction of Diesters to Hydrocarbons .— The reduction of diethyl 3,6-dimethylidithalate by lithium aluminum hydride gave 3,6-dimethyl-c- xylylenol in yields of 87-903. Much larger quantities of ester than 48 E, Buchta and G. Loew, Ann., 123 (1956). -6ii- anhydride could be reduced at one time and the use o f a Soxhlet outfit iras eliminated. When this method was applied to diethyl 3-methylphthalate and the esters of the other anliydrides prepared in this investigation, very good results also were obtained. The Birch reduction of 3,6-\limethyl-o-zylylenol proved to be an excellent method of converting the diol to prehnitene. For example; when one mole of the compound and two moles of methanol (in 1 liter of ether) was added to a solution of four moles of sodium in liquid ammonia, reduction took place rapidly and cleanly to afford the hydrocarbon in $0^ yield. Other runs of one mole or larger con sistently gave yields of product of b8-9C^# When this reduction method was applied to the other diols prepared, similar excellent results were obtained. No evidence for nuclear reduction by sodium was found in any case (aqueous potassium permanganate test)* Reduction of Esters by Sodium and Methanol in Liquid Ammonia.— Only unsaturated products which could not be separated by distillation were obtained idien one mole of diethyl or dibutyl 3,6-dimethyl- phthalate was dissolved in 12 moles of methanol and added to a solu tion of 12 moles of sodium in liquid ammonia. Ethyl phthalate itself gave only unidentifiable products and no o-xylene. Preparation of Esters.— No difficulty was encountered in the preparation of a majority of the esters needed for this research program. A typical procedure using sulfuric acid as the catalyst (in -65- larger qxiantltifjs than ia normally required) gave good yields of esters -rrhen -prater was continuously removed (as aseotrope) for two or three da^-s. Tivo anhydrides (3-methyl-6-ethylphthaL ic and 3-methyl- 6-n-propyIphthalic ) would not respond to this method and had to be prepared in another way. In lylUjj Geissman and Morris'^® prepared dimethyl naphthalate in 89$ yield by dissol-vlng naphthalic anhydride in methanol containing -two equivalents of potassium hydroxide and adding, at a rate of Uîl, six more equivalents of base in methanol and eight moles of dimethyl sulfate. The potassium methoxysulfate was filtered off and the methanol solution v/as cooled by an ice-salt mixture in order to obtain the ester. When applied to the difficult anhydrides and 3j 6-dimethylphthalio anhydride, the dimethyl esters were obtained in yields of 60-72$. A careful study of this procedure was not made in an attempt to improve yields. Conclusion.— The hydrocarbons prepared required no purification other than a simple vacuum distillation. Purities of distilled products ranged from 90$ to 99.9$# Although only moderate quantities of most of the hydrocarbons were prepared by the over-all scheme, it is easily adaptable to large scale production. The American Petroleum Institute Research Project US later used the synthetic route to prepare two and one-half liters of prehnitene. 49 T.A. Geissman and L. Morris, J. Am. Chem. Soc., 66, 71b (19hl4 ). «66“ Accurate boiling points, freezing points, densities, purities, and refractive indices were detennined for all hydrocarbons pre pared^ elemental analyses (carbon and hydrogen) of all near hydro- cai'bons v/cre obtained. The intermediate diols v/ere characterized by melting points (or boiling point range, in a single case) and elemental analyses (carbon and hydrogen). Diesters >vere^‘likewise characterized. EXPERIMENTAL Determination of Physical Properties and Purities of Hydrocarbon Products Cryoscopic Determinations.— The apparatus trhich was used for the determination of freezing points, melting points, and mixed melting points was essentially that which has been described by Glasgow, Streiff and Rossini^. Temperatures were measured by a platinum re sistance thermometer (calibrated by the National Bureau of Standards), in connection with a Mueller resistance bridge (Leeds and Northrup, Type G-2). With this apparatus temperature differences of less than 0 .0 1 * can be measured accurately. The purity of each caapound for trtiich no cryoscopic constant was available was determined by essentially the sane method as that which was described by Mair, Glasgow, and Rossini*: the theoretical freez ing point of 100 percent pure material (Tfo) was determined from the freezing point or melting point of the substance and the midpoint temperature on a time-temperature freezing curve* At the midpoint, one-half of the compound is crystallized and the midpoint temperature represents the freezing point of the compound containing exactly twice the amount of impurity than it does at the actual freezing or melting point. The true freezing point, Tfo, then can be calculated 1 A.R. Glasgow, A.J* Streiff, and F.D* Rossini, J. Res. Natl. Bur, Stand., 355 (19ii5). 2 B.J. Mair, A.R. Glasgow,and F.D. Rossini, J. Res. Natl. Bur. Stand*, 591 (19U1). -67— "68— from the fact that the difference between the true freezing point and the observed freezing or melting point is the same as the difference between the observed freezing or melting point and the midpoint, A value for the cryoscopic constant A (mole fraction per degree) then was determined experimentally by adding a weighed amount of suitable impurity (iso-octane) to a known weight of the material being in vestigated and measuring the resultant lowering of the melting point. Upon obtaining the values of Tfo, A, and Tm (the melting point), the purity (p) of the compound was calculated, using the formula Boiling Points«^The boiling points of the hydrocarbons vhich were prepared during this investigation were determined in a modified Cottrell apparatus (described by Qulggle, Tongberg, and Fenske^) ^Aiich was attached to a barostat that maintained a constant pressure of 760 am. of mercury* Nitrogen was used in this system. The temperatures were measured with the same platinum resistance ther mometer and resistance bridge which was used for the cryoscopic de terminations • Most of the cryoscopic and boiling point determinations were carried out by Mr. Kenneth Davis of the American Petroleum Institute Research Project U5* 3 D* Qoiggle, C.O* Tongberg, and U.R. Fenske, Ind. Eng. Chem., Anal. Ed., ^ U66 (193b). —69— Refractive Indices » — Refrac tive indices were determined at 20" with a Valentine Precision Refractometer (manufactured by the Industro-Scientific Company) which was connected to an electroni cally controlled constant temperature bath. Densities. — The densities of the aromatic hydrocarbons were determined at 20" with the use of a 20-ral. pycnometer which had been calibrated accurately with standard iso-octane. A constant tempera ture bath controlled the temperature to i 0 .0 2 " and all weights were corrected to weights In vacuo before the densities were calculated. Special Apparatus 1, The distillation columns which were used in the course of this investigation are designated as column A, B, or C in connection with each distillation which was carried out. These are described below: column A 0.8 x 50 cm,, packed with 3/32-in. Pyrex glass helices, fitted for distillations at atmospheric pressure only, with a probable maximum efficiency of 15 plates. column B 1.3 x 50 cm., packed with l/S-ln* Pyrex glass helices, fitted for vacuum or atmospheric pressure distil lations, with a probable maximum efficiency of 15 plates when operated at atmospheric pressure. -70- column C 1.3 X 100 cm., packed with 3/l6-in. Pyrex glass helices, fitted for vacuum or atmospheric pressure distillations, with a probable maximum efficiency of 15 plates when operated at atmospheric pressure. Other columns which were used occasionally are described in con nection with the particular distillations in which they served. 2. Neutral equivalents of compounds were determined with a Beckmann model H-2 pH meter, in connection with a Type 225-A glass electrode and a Type 256-A fiber-type reference electrode. 3* Melting points of all solids were determined with the use of an electrically heated. Internally illuminated aluminum block into Trtiich could be inserted several sample-filled capillary tubes and a thermmmeter. All melting points are uncorrected. Production of 2-Alkylfurans and 2-Methyl-5-^^ kylfurans 2-Metbylfuran.— One gallon of commercial 2-methylfuran (du Pont) first was distilled carefully through a 1.8 x 150 cm. column packed with l/B-in. I ^ e x glass helices, but the pure fur an was not obtained. The refractive indices of all of the fractions which were collected were considerably lower than the value reported for 2-methylfuran by the American Petroleum Institute Research Project 1.U332). A second distillation of the combined fractions (through an eight- foot, vacuum-jacketed silvered column packed with l/8-in. lyrex 4 l5th XnzmaïrReport, American Petroleum Institute Research Project U5, 1956, Cumulative Table. -71- glas3 helices) at about a $0-plate efficiency gave two main fractions. The first fraction was good 2-methylfuran, b.p. 6L-6ü«l'/7Ü7 mm., 1.1^332; 2.7 kg. were purified in this manner. The second frac tion (boo g. ) distilled at 66-66 .l*/7b7 mm#, and had a refractive index of 1.U150-1.U155 (n§°). Its infra-red spectrum contained a strong band at 1075 cm#"^ (C-O-C stretching frequency of a cyclic ether), a strong band at ?15 cm.1 (CH out-of-plane deformation of a vinyl 0 = 0), a medium band at l650 cm."^ (C = C stretching vibra tion) and a strong band at 3002 cm."^ (CH stretching vibration of a 0 = 0). From this Information the compound was tentatively con cluded to be a 2-methyldlhydrofuran or a methylenedihydrofuran. The physical properties of all of these structures (except one) were compared with those of the unknown compound, but were found to be considerably different* One possible structure, 2-methyl-2,3-di- hydro furan, was unreported in the literature. It is not unlikely that the fraction distilling at 66-6 6 *1 " in the above distillation was the 2, 3-dihydrofuran * 2-Fthylftaran. — Two procedures were investigated in an effort to prepare this furan in a reasonable yield. The first procedure, that of Paul®, proved to be unsatisfactory. One of several runs is des cribed below: g R* Paul, Conpt* rend*. 200. 11*81 (193^), -72- a, 2-Furylmethylcarbinol was prepared essentially according to the method of Yurev and Gragerov®: Into a 5-1» round-bottomed, three-necked flask equipped with a motor-driven Hershberg-type stirrer, an addition funnel, and a Liebig condenser (provided with a calcium chloride tube) was placed Ü2.0 g. (3*ii moles) of magnesium turnings. The flask was flushed with nitrogen, 1 liter of dry ether was introduced, and $0 ml. of a solution of h$h g» (3*2 moles) of laethyl iodide (E*K.) in 1 liter of dry ether was added. Tfhen the reaction began, the flask was sur rounded by an ice-water bath and the remainder of the methyl iodide solution was added dropwisci with stirring, through the addition funnel. Upon completion of the addition, the cooling bath was re moved and the mixture was stirred for three hours. Titration of a hydrolyzed 10-ml. aliquot sample of the solution indicated that 3 moles of Grignard reagent had been formed. The solution then was cooled to -15** by a Dry Ice-isopropanol bath and a solution of 1Ü2 g. (2 moles) of freshly distilled furfural in 300 ml. of dry ether was added dropwise. The cooling bath was removed and the reaction mixture was wanned on a steam bath for 0.5 hr. The flask was surrounded by an ice-water bath and sufficient water was added through the addition funnel to cause the magnesium salts to settle out in the form of small pellets. The ether solution was decanted. 6 Y.K. Yurev and I.P. Gragerov, Zhur. Obshchei Khim., 19, 72h (19h9); G . A . , ^ 1092 (1950). -73- the pellets were washed with trso l^O-ail. portions of ether, and the combined material (solution and extracts) was dried over anhydrous sodium sulfate. The ether was removed and the product was distilled under reduced pressure (in the presence of 2 ml. of aniline) through column B to give 3Ji6 g. (1.3 moles, 65%) of 2-furylmethylcarbinol, b.p. 7ô-79“/23 mm., ng° l.J^SlO (reported®: b.p. 76-77“/23 mm., n^° 1 .1:8 0 0 ). b. Two hundred and twenty-four grams (2 moles) of 2-furylmethyl- carbinol was dehydrated over aluminum oxide in the following manner; the alcohol was introduced at a rate of one drop every three seconds (by a metering device) into a 1.8 x 110 cm. Pyrex glass tube which was packed with pelleted catalyst and was maintained at h50* by a vertical tube furnace. The effluent vapors, upon passing through a condenser attached to the lower end of the tube, condensed and the liquid products collected in a Dry Ice cooled receiver. The organic material was isolated, dried over anhydrous sodium sulfate, and distilled at atmospfieric pressure through column A to give 66 g. (0.7 mole, 35%) of 2-vinylfuran, b.p. 96.8-98.8*/7h6 mm., 1.1:986 (reported'': b.p. 99-100*/760 mm., 1.1:992). Hïie inside surface of the tube and the aluminum oxide pellets were found to be coated with a considerable amount of dark resin. A dehydration run carried out at 350" also gave much polymer. V C. Moureu, C. Dufraisse, and J.R, Johnson, Annales de Chemie Fiol, 2, 20 (1927). -7ü- A useful, rapid method of dehydrating 2-furylmethylcarbinol was worked out by the present author. It is exemplified here by the run which afforded the best yield of 2-vlnylfuran: One hundred and ten grams (1 mole) of the alcohol and a small crystal of iodine about the size of a pin head were placed in a 250-ml. Claisen flask fitted for a vacuum distillation. A steam bath was placed beneath the flask and the single receiver was surrounded by a Dry Ice-acetone bath; between the system and the vacuum pump was included the usual Dry Ice-acetone cooled trap. The pressure within the system then was reduced to 1^0 nun. and heat was applied to the flask in an ever-increasing amount as the dehydration took place. When only a dark, viscous tar remained in the flask the steam bath was removed, the flask was allowed to cool, and atmospheric pressure was restored to the system. The receiver and trap contents were com bined and the organic layer was isolated, dried over anhydrous sodium sulfate, and distilled under atmospheric pressure through column A to give 62 g. (0.65 mole, 65$) of 2-vinylfuran, b.p. o8.1-5>8.3*/7U5 mm., ngo 1.4883. c. Fifty grams (0.53 moles) of 2-vinylfuran (in 100 ml. of ether) was hydrogenated in a Parr-tj’pe bomb at 25* under a hydrogen pres sure of 3*5 atm., and in the presence of 5 g* of platinum oxide (pre pared according to Adams, Voorhees and Shriner®;j, After the the oretical amount of hydrogen had been absorbed (24 hr s. ) the catalyst a R. Adams, V. Voorhees and R.L. Shriner, Organic Syntheses. Cq H. Vol. I., John Wiley and Sons, Inc., New York, N.I., 1941, p. 463. -75- ras filtered, the ether was removed and the product was distilled under atmospheric pressure through column A to give 30.6 g. (0.32 mole, 60^) of 2-ethylfuran, b.p. 89.8-B9.9°/7L5 mm., 1.^391 (reported^: b.p. 92-93"/760 mm., l.f4^6 6 ). The temperature then rapidly rose to 98* and the remaining material distilled between 98 * and 103.5** A mixture of 2-cthyltetrahydrofuran and unreduced 2-vlnylfuran was assumed to be present; however, the efficiaicy of column A was not high enough to effect the separation of these sub stances. ■pihen a palladiTim-on-charcoal catalyst was substituted for platinum oxide in a subsequent run, the result was essentially the S32QS # At a later period of this research program, an excellent method of reducing furylolefins to alkylfur ans (without effecting the re duction of the furan ring) was discovered by the present author. Its application to 2-vinylfuran is exemplified below; A 3-1* round-bottomed, three-necked flask was equipped with a motor-driven Hershberg-type stirrer, an addition funnel, and a special Dry Ice-acetone cooled condenser (described and illustrated by Greenlee®). For the purpose of insulation, asbestos fiber was packed about the lower half of the flask. One liter of liquid ammonia then was fed into the vessel from the liquid phase of a commercial e K.W* Greenlee and A.L. Henne, Inorganic Syntheses^ Vol. II, McGraw-Hill Book Co., Inc., 19^6, pp. 130-131* «76“ tank, Ijô.l g» (2.1 moles) of freshly cut sodium iras added in approximately ?-g. pieces, and stirring was commenced. 'When the sodium had dissolved (about one-half hour), a solution of 9U g« (1 mole) of 2-vinylfuran in ^00 ml. of dry ether was added dropwise through the addition funnel in one hour. When the addition was completed the funnel was removed, and the sodium amide and excess sodium were neutralized with IlU g. (2,1 moles) of powdered ammonium chloride* This solid reagent was added from a special flask (an Erlenmeyer with neck bent 90* from normal) which was attached to the reaction flask with a section of Gooch tubing. The funnel then was replaced and through it was added cold water in a steady stream until ammonia no longer refluxed. Stirring was discontinued, the aqueous layer was siphoned off and discarded, and the ether solution was washed with water, dried over anhydrous sodium sulfate and dis tilled through column A to give 92 g. (0.96 mole, 96^) of 2-ethyl furan, b.p. 90.1-90.2*/7W mm., n£° 1.L396. The second of the two procedures idiich were examined in the search for a satisfactory route to 2-ethylfuran proved to be an ex cellent one* One of the runs is described below; a. According to the directions of Held and Levine^®, 3liO g. moles) of freshly distilled furan and 6l^ g. (5*75 moles) of acetic anhydride (Carbide and Carbon) were mixed in a 3-l« round-bottomed, three-necked flask equipped with a motor-driven glass stirring rod xo J.V. Held and R. Levine, J, Org* Chen., 1 ^ U09 (19L8), -77- (to which was attached a Teflon blade), a thermometer, and a four- foot Liebig condenser from which was led a tube to a Dry Ice cooled trap. By means of a salt-ice mixture tlie flask contents were cooled to vigorous stirring then was consnenced, the thermometer was removed, and 70 g. of purified boron trifluoride-diethyl ether was added all at once. A rubber stopper quickly was placed in the neck through w M c b the catalyst had been added. A vigorous reaction took place which generated sufficient heat to reflux acetic acid; how ever, examination of the Dry Ice cooled trap revealed that only a small amount of furan had passed out of the condenser. Mien the reaction mixture had cooled to room temperature, the cooling bath was removed, stirring was maintained for one-half hour, and the dark-blue liquid was poured into a U-1 * separatory funnel which contained 2 liters of water and 1 liter of chloroform. The funnel was shaken vigorously until the color of the organic phase changed from dark blue to dark red, after which the organic layer was removed and the aqueous layer was extracted with two 1^0 -ml# portions of chloroform. The combined material (layer and extracts) was washed with 1 liter of a saturated sodium carbonate solution and dried by percolation through a column of anhydrous sodium sulfate. The chloro form was removed and the product was distilled under reduced pres sure through column B to give 275 g* (2.5 moles; $0%, based upon furan) of 2-furyl methyl ketone, b.p. 6l.^-62 .5 */l2 mm., n|° 1 .5o62 , -78- 2,l|-diiiltrophei3ylhydrazone m.p. 221*5-222.5“ (reported^b.p. 62- 63"/12 mm., ng° 1*50167 , 2,U-dinitrophenylhydrazone m.p. 219-220" ). One other acétylation procedure^* was tried, jn which a boron trifluoride-acetic acid catalyst was used. Since this method gave yields of 2-furyl methyl ketone similar to those obtained when boron trifluoride-diethyl ether was used, it will not be exemplified here. b. The reduction of 2-furyl methyl ketone to 2-ethylfuran was accomplished by a modified Wolff-Kishner procedure which had been developed by the API Research Project U5^^* This procedure proved to give excellent yields of products in all reductions of ketones which were carried out by the present author. A large run is described below: One thousand and sixty-nine grams (6,6 moles) of dihydrazine sulfate, 528 g. (13.2 moles) of sodium hydroxide pellets, and 2 liters of diethylene glycol were placed in a 5-1 » round-bottomed, three necked flask equipped with a motor-driven stainless-steel stirrer of the Hershberg type, a thermometer, a large Liebig condenser, and an electrically heated mantle. The mixture was stirred and heated until the dihydrazine sulfate had melted and a vigorous reaction had begun (at 115* )j the temperature rapidly rose to 130“ during this reaction. Iff probable value 1 .5 0 6 1 7 * 11 A*P* Dunlop and F.N* Peters, The Furans. Reinhold Publishing Corp., New York, N.T., 1953, p. 1:30. 12 R. Levine, J.V. Heid, and M.W. Farrar, J. Am. Chem. Soc., 71, 120? (191:9). 13 18th Annual Report, American Petroleum Institute Research Project 1:5, 1956, p. 56. -79- Heat was removed until the temperature dropped to 110", and 721) g. (6.58 moles} of 2-furyl methyl ketone was added rapidly. The mixture was refluxed for three hours, cooled to 100", and itOO g. (10 moles) of sodium hydroxide was added. The condenser then was removed and replaced by a wide-bore (i*d* 25 mm. ) glass tube which was led to an ice-water cooled 5-1 . two-necked, round-bottomed flask equipped with a four-foot Liebig condenser from vdiich a tube was led to a Dry Ice cooled trap* The product was distilled from the reactor while vig orous stirring was maintained. Most of the 2-ethylfuran condensed in the receiver along with a considerable amount of water, but some was swept into the Dry Ice cooled trap by escaping nitrogen. In about two hours distillation had nearly stopped j 200 ml. of water then was run slowly into the reaction flask (through a dropping funnel) to sweep any remaining product into -die receiver, lïhen distillation again ceased (in about an hour) the reaction flask was allowed to cool, then was disconnected and washed out with hot water before the residue could solidify. The organic material from the condensing system was added to that from the trap and the combined material was washed with a 2% hydro chloric acid solution, dried by percolation through a column of anhydrous sodium sulfate and distilled through column B to give h92 g. (5.1 moles, 77,2%) of 2-ethylfuran, b.p. 9O-90.2"/750 mm., n|o 1.W95. -8o~ A total of 1316 g. of 2-ethylfuran was prepared by the three methods described. 2-n-Propylfuran >■— This compound was prepared by the same two- step procedure which was used in the preparation of 2-ethylfuran (pp. 76 to 7 9 ): 1 ) furan was propionylated with propionic anhydride in the presence of a boron trifluorlde-diethyl ether catalyst according to the procedure of Heid and Levine^®, 2) the resultant ethyl 2-furyl ketone was reduced to 2-n-propylfuran by the modified Wolff-Kishner method^. Therefore, only a brief account of a propionylation run and a reduction run is given: a. To a cooled and rapidly stirred mixture of UOÜ g. (6 moles) of freshly distilled furan and Ü97 g. (6.9 moles) of propionic an hydride (Carbide and Carbon) was added 8U g» of purified boron tri- fluoride-diethyl ether all at once. The temperature of the reaction mixture rcpidly reached a maximum of 120®, then dropped to 30“ in about 15 minutes. Cooling was discontinued, but stirring was main tained for one-half hour longer. The vessel contents then were hydrolyzed and worked up for product. Distillation under reduced pressure through column B gave U89 g. (3.9 moles, 66$) o f ethyl 2-furyl ketone, b.p. 63-6h*/6 mm., m.p. 30-31“, semicarbazone m.p. 188-188.5“ (reported; b.p* 6l-63"/6 mm., semicarbazone m.p. I86- I87’^°j m.p. 28-30“^“ ). 14 J, Ramonczai and L. Vargha, J. Am. Chem. Soc., %2, 2737 (1?50). -81- b. Four hundred and eighty-nine grams (3.9 moles) of ethyl 2-furyl ketone was added to a stirred solution of ti moles of hydrazine |~prepared in situ from 6&8 g. (U moles) of dihydrazine sulfate and 320 g« (8 moles) of sodium hydroxide pelletJj in 1,5 liters of diethylene glycol. The mixture was refluxed for three hours and cooled to 100*. Two hundred and forty grams (6 moles) of sodium hydroxide was added and the furan, hydrazine, and water were distilled from the reaction vessel. Two hundred milliliters of water then was added in order to sweep any remaining material into the re ceiver, and when the distillation again ceased the receiver and trap contents were combined and worked up for product. Distillation of this product through column B gave 363 g. (3.3 moles, 85%) of 2-n-propylfuran, b.p. U3»5"lli+’/739 mm., ng<* l*hlil? (reported^®; b.p. m .5-115.^V750 mm., ng° l.iU,5 5 ). A total of 1001 g. of 2-n-propylfuran was prepared by the above two-step procedure. 2,5-Diaethylfuran.— One gallon of commercial 2,5-dime thyIfur an (Carbide and Carbon) was distilled slowly throu^ a 1.8 x 150 cm. column packed with l/8-in. î ^ e x glass helices. However, all of the fractions which were collected distilled at lower temperatures and exhibited considerably lower refractive indices than those which have been reported** for the pure compound (b.p. 9U.l*/760 mm., n^° IS N.I. ^u ï t j n , E.V. Shemaetina, and E.D. Cherkasova, J. Gen. Chem. Ü.S.S.H., 8, 67li (1938); C.A. I3I6 (1939). -02- l.ltiilO ). Therefore, these fractions were combined and subjected to a slow distillation through a column of higher efficiency (5>0 plates, described on p. 10 ), Two hundred and fifty grams of an unidentified compound distilled at 91*5“/i k ^ mm. (n^° 1 .^2 3 2 ), followed by a small amount of intermediate material, then 3 kg. of 2,5“diaethylfuran, b.p. 9[i-9ii.l*/7lih mm«, 1 .14396-1 .Wi00, Although the value of the refractive index of this material was somewhat low, its reactivity with raaleic anhydride was not impaired. Excellent yields of adduct were obtained. 2-Methyl-^-ethvlfuran.— The preparation of this compound was accomplished by acetylating 2-aethylfuran according to the procedure of Farrar and Levine^®, and reducing the resultant ketone by a modified Wolff-Kishner method^The above investigators^ procedure was modified only to the minor extent that a lower initial cooling of reactants was applied and the formed ketone was rapidly strip-dis tilled prior to its final distillation. Otherwise, both the acétyla tion and the reduction steps were carried out in the same manner as in the preparation of 2-ethylfuran (pp. 76 to 7 9 ). Therefore, they will not be described in detail. A brief outline of the acétylation run affording the best yield of ketone and a typical reduction run is given below: 18 U,lf. Farrar and R, Levine, J. Am* Chem. Soc., %2, 3695 (1950). -83" a. To a rapidly stirred mixture of U92 g. (6 moles) of purified 2-aethylfuran and lOh g. (6.9 moles) of acetic anhydride (Carbide and Carbon), cooled to -10* by means of a Dry Ice-isopropanol bath, was added Bit g. of purified boron trifluoride-diethyl ether all at once. TTithin 30 sec. the temperature of the reaction mixture reached its maximum, 113* « When the temperature had fallen to 30*, the cooling bath was removed and the mixture was stirred for one-half hour longer. The vessel contents (dark blue in color) then were hydrolyzed and worked up for product. The product was distilled very rapidly under low pressure (3 mm. ). The distillate which was obtained (280 g. ) then was distilled carefully under reduced pressure through column B to give 260 g. (2,1 moles, 33^) of 3 ”0 ©thyl-2-furyl methyl ketone, b.p. 72.6-73*/8 mm., ng® 1*5135, semicarbazone m.p. 182-183* (re ported^®: b.p. 71-73*/8 mm., semicarbazone m.p. 190,5 "191#5 * )« b. Three hundred and ninety grams (3.lit moles) of 5-methyl- 2-furyl methyl ketone was added to a solution of 6 .U moles of hydrazine j^epared ir situ from 318 g. (3.2 moles) of dihydrazine sulfate and 256 g. (6 .U moles) of sodium hydroxide pellet^ in 1,5 liters of diethylene glycol and the reaction mixture was refluxed for 3 hrs. After allowing the flask contents to cool to 100", 192 g. (U.8 moles) of sodium hydroxide was added and the furan, hydrazine, and water were distilled from the reaction vessel. Two hundred milliliters of water then was added in order to sweep any remaining material into the receiver, and when the distillation again ceased “81j“ the receiver and trap contents were combined and worked up for product. Distillation of this product through column B gave 27? g« (2.5 moles, 8056) of 2-methyl-5-ethylfuran, b.p. 11^.2-ll?.?"/73? mm., ngo l.li!i73 (reportedly: b.p. 1 16-U8V7U2 mm., l.JiU.'j). A total of 680 g. of 2-nethyl-5“ethylfuran was prepared. 2-Me thy l-?-g-propy Ifur an.— This di substituted furan also was prepared by a procedure which incorporated an acylation step and a reduction step. The propionylation of 2-methylfuran as carried out by the present author differed little from the method employed by Farrar and Levine^®. These investigators found that it was not necessary to cool the reactants (0.5 mole 2-methylfuran and 0*58 mole propionic anhydride) before the catalyst (boron trifluoride-diethyl ether) was added because the reaction was not particularly vigorous. However, in the presently described research, where the quantities of reactants were increased six-fold, the reaction proved to be very exothermic and unless moderated somewhat, gave rise to 85$ or more polymer. "Rhen warm water was used to effect this moderation of temperature the yield of desired ketone was increased remarkably. A preliminary rapid vacuum distillation of the ketone from polymer formed during this reaction also increased the yield of product. Other than the modifications which were just described, both the acylation and reduction steps were carried out in the same manner as those vdiich were used to prepare 2-ethylfuran (pp. 76 to 19), and -ü$- therefore they will not be presented in detail. The propionylation run which gave the highest yield of ketone and a reduction run using the modified Wolff-Kishner method^^ are outlined below: a. To a rapidly stirred mixture of 2&6 g, (3 moles) of purified 2-methyl furan and g. (3 »h5 moles) of freshly distilled propionic anhydride, contained in a flask which was surrounded by a warm water bath, was added h2 g. of boron trifluoride-diethyl ether all at once* A very exothermic reaction ensued which lasted for three or four minutes, much longer than the other acylation reactions which have been described. Tbs temperature of the flask contents reached a maxi mum 110*. The bath was removed and when the temperature of the reac tion mixture had diminished to U0 °, the vessel contents (dark blue in color) were hydrolyzed and worked up for product. This product was very rapidly distilled under low pressure (2 mm.). The distillate (2liO g. ) which was obtained then was distilled slowly under reduced pressure through column B to give 210 g. (1.^ moles, ^056) of ethyl S-methyl-2-furyl ketone, b.p. 77.2-77* ^ * A o mm., n|° 1 .^060, semi carbazone m.p. 13$.$-1S6 " (reported^®; b.p. 69.^-70“/h*5 mm., semi carbazone m.p. 156-1^7*)• b. Four hundred and twenty grains (3.1 moles) of ettqrl ^-methyl- 2-furyl ketone was added to a solution of 6.2 moles of hydrazine l^repared in situ from U86 g. (3.1 moles) of dihydrazine sulfate and 260 g. (6.2 moles) of sodium hydroxide pellet^ in 1.5 liters of ••86*" diethylene glycol and the mixture was refluxed for 3 hrs. After allowing the flask contents to cool to 100*, 188 g. (i|,7 moles) of sodium hydroxide was added and the furan, hydrazine, and water were distilled from the reaction vessel. Two hundred and fifty milliliters of water then was added in order to sweep any remaining material into the receiver, and when the distillation again ceased tho combined re ceiver and trap contents were worked up for product. Distillation of this product through column B gave 302 g. (2.1i moles, 80%) of 2-methyl- 5-n-propylfuran, b.p. 137*^-137"7mm», n^° l.UUdl (reported^: b.p. 138V773 mm., ng» l.l^h82). A total of 7U7 g* of 2-methyl-^-n-propylfuran was prepared. 2-l80propylfuran.w-Considerable difficulty was encountered when efforts were made to prepare this compound by existing procedures. The first method which was examined, that of Oilman and Calloway^s, had to be rejected in its entirety after numerous experiments had failed to produce a suitable means of carrying out the initial step (a Friedel-Crafts Isopropylation of methyl 2-furoate). A description of a typical unsuccessful attempt is given: One and one-half liters of carbon disulfide (distilled from I^osphorus pentoxide) and 270 g. (2 moles) of anhydrous aluminum chloride (comm.) were placed in a 3-1 . round-bottomed, three-necked flask equipped with a motoi*-driven Hershberg-type stirrer, a 17 11. Fetizon, Compt. rend., 236. k99 (1953)» 16 H. Gilman and N. Calloway, J. Jim. Chem. Soc., IH9J (1933), -Ü7- thermometer, an additional funnel, and a small Liebig condenser (into the top of which was placed a loose plug of glass wool). The flask contents were cooled to -5" by means of a Dry Ice-isopropanol bath, stirring was commenced, and a mixture of 126 g* (1 mole) of methyl 2-furoate jprepared from 2-furoic acid (Q.O.)^^ and 79 g« (1 mole) of is opr 0 ^ 1 chloride (EK White Label) was added dropwise through the addition funnel* The temperature of -5"C. was maintained throughout the addition, after which the cooling bath was removed, stirring was discontinued, and the reaction mixture was allowed to stand for 12 hrs. It then was poured slowly over about 3 kg, of crushed ice, with vigorous stirring* When the ice had melted the organic layer was siphoned and the aqueous layer was extracted with two 200-ml. portions of carbon disulfide. The combined material (layer and extracts) was washed with ^00 ml. of 10$ hydrochloric acid and dried by percolation through a column of anhydrous sodium sulfate. The solvent was re moved and the product was distilled under reduced pressure through column B* Distillation Fraction Temp. Pres5ure(mm* ) Rate ( dps ./sec. ) Wt*(g. ) Remarks 1 79-81" 20 1/20 20 2 81-83* 20 1/20 12 3 83-83.8" 20 1/20 37 ngo 1 . 4 8 8 1 1 » a.0* Clinton and S*C. Laskowski, J, Am. Chem. Soc*, JO, 3135 (1948). -ÜÜ- Fraction Temp. PreasureÇmm.) Rate (dps./sec.) Wt»(g» ) Remarks U 63.8-100* 20 1/15 5 ngo 1.1)898 yellow oil with evil odor 5 100-110* 20 1/15 20 ngo 1.1)952 orange oil with evil odor 6 110-130* 20 1/15 10 orange oil with evil odor 7 130-11)0* 20 1/15 10 orange oil with evil odor residue 100 deep-red, viscous oil with evil odor The boiling range and refractive index of pure methyl 5-iso- propyl-2-furoate have been reported^» as 110-112"/20 mm. and ng® 1*1)851. Thus, it can be seen that pure ester was not obtained in the above distillation* After a number of unsuccessful experiments had been carried out using carbon disulfide as solvent, it was concluded that this solvent had been taking part in the Friedel-Crafts reaction. Therefore, two other solvents were tried (nitromethane and sym-tetrachloroethane ). Only polymer resulted when ni tromethane was employed, but a certain measure of success was realized (in one of four experiments) when the -y?- isopropylation was carried out in the chlorinated hydrocarbon. This experiment is not presented in detail here since the procedure which wao used was the same as the one described on p. 8 6 , A mixture of 359 g* (2*9 moles) of methyl 2-furoate and 237 g. (2.9 moles) of isopropyl chloride was added slowly to a stirred, cool ed suspension of 5U0 g. (L.O g. (h-O moles) of anhydrous aluminum chloride in 2 liters of freshly distilled sym-tetrachloroethane. The reaction mixture then was allowed to stand for 12 hrs, at room temperature, after which it was hydrolysed and worked up in the manner previously described. Distillation of this product under re duced pressure through column B yielded 96 g. (0*57 mole, 19*5$) of methyl S-isopropyl-a-furcate, b.p, 11^,3-115*0"/20 mm,, 1.L851- l,Ud55* The residue of this distillation was a black, viscous oil; it amounted to 220 g. The ester which was obtained was not con verted to 2-lsopropylfuran, The second procedure for synthesizing 2-isopropylfuran, which was examined by the present author, was a three-step one which had been reported by Reichstein®®. These three steps are considered separately below wJ.th respect to their preparative value; several changes in these steps were found to be necessary (or found to give better yields of pi*oducts) by the present author and one other author*^. 20 T, Reiciistein, Helv, ChimJ Acta, 1 ^ 1118 (1932). 21 G,B. Bachmann and L.V, Reisey, J, Am. Chem. Soc., Jl, 1985 (19Ü9). -90- a. In Reichstein^s original process, dlmethyl-2-furylcarbinol was obtained by a typical Grignard procedure in which 2-furyl methyl ketone and méthylmagnésium iodide were the reactants; the yield of product was good. At a later date, however, Bachman and Heisey^^ showed that an equivalent yield of dimethyl-2-furyIcarbinol could be prepared from the more easily obtained methyl 2-furoate, and méthylmagnésium iodide. Therefore, in the presently described re search, the original method of Reichstein was used only in two small experimental runs. For the preparation of the remainder of the de sired amount of alcohol, the second method was employed; for pract ical purposes méthylmagnésium chloride was substituted for méthyl magnésium iodide. The largest run is described here: A large-capacity, water-jacketed, Monel Metal lined Grignard tank, specially constructed for the API Research Project hSt was used in this preparation. This removable unit is fitted to a large metal plate, part of a permanent setup which consists of a motor-driven, bearing-sealed, Monel Metal stirrer; a metal-jacketed, copper-tube condenser; and a protective blow-out system containing a large over flow tank. Three hundred and eighty-nine grams (16 moles) of magnesium metal turnings was introduced into the reactor (through a loading hole in the above-mentioned plate), the system was flushed thoroughly with nitrogen, and the magnesium was covered with 2 liters of dry ether. Fifty milliliters of methyl iodide was added and when the reaction began, a gas-inlet tube (which extended below the -91- surface of the liquid) was fitted into place, stirring was com menced, and methyl chloride was introduced at a steady rate from a commercial tank. The vessel was cooled when the reaction became too vigorous by passing cold water through its jacket; three more liters of ether were added during the course of the preparation in order to prevent the reagent solution from becoming viscous. At intervals the methyl chloride addition was stopped, stirring was discontinued, and a sample of material was dipped up from the bottom of the reactor vessel. When magnesium was no longer obtained in a sample (3 hrs. ), the reaction was judged to be complete. A lO^nl. aliquot portion then was hydrolyzed and titrated for base (by overtitrating with a standard hydrochloric acid solution then backtitrating with a standard sodium hydroad.de solution). From the value obtained from this anal ysis, the total number of moles of reagent was calculated to be 19 .2 . With stirring and continued cooling, a solution of 882 g. (7 moles) of methyl 2-furoate in 2.9 liters of dry ether then was added at a fairly rapid rate through a 2-1. addition funnel. When the addition was com pleted, warm water was passed through the reactor jacket until a steady reflux was maintained. After 2 hrs. the tank was cooled to 10-15" and a sufficient amount of water (l.U liters) was added through the funnel to cause the magnesium salts to settle as pellets. The ether solution was siphoned under a slight positive air pressure and the pellets were washed with 2 900-ml. portions of ether. The com bined material (solution and washings) was dried by percolation -92- through a column of anhydrous sulfate, the solvent was removed, and the product was distilled under reduced pressure through column G to give ^Oli g. (Il moles, ^7%) of dimethyl-2-furylcarbinol, b.p. 66 "/l$ mm., ngo 1.U732 (reported®^; b.p. 71-72°/l^ mm., ng® l.h700). The Dry Ice cooled trap yielded a considerable amount of material, which was distilled through column B to give 108 g. (1 mole, li^ ) of 2-1 sopropenylfuran, b.p, 12$.2-12$.^"/7Ü7 mm., np° l.*f996 (reported®^! b.p. 12^-126"/atm., ng® 1.U966)# Therefore, the total yield of product from the Grignard reaction was 71^. b. The second step of Reichstein*s procedure®®, the dehydration of dimethy1 -2-furylcarbinol with acetic anhydride (at the reflux temperature), could not be carried out to give a creditable yield of the olefin in the presently described research. Reichstein, himself, was only able to realize a 23% yield. However, Bachmann and Heisey^i reportedly improved the method by using anhydrous sodium acetate in conjunction with acetic anhydride. "When these investigators* pro cedure was employed by the present author, good yields of 2-iso- propenylfuran were obtained. The larger of two runs is described below; Four hundred grams (3.17 moles) of dimethy 1-2-furylcarbinol, 326 g. (3*2 moles) of acetic anhydride (Carbide and Carbon), and 112 g. (1.6 moles) of anhydrous sodium acetate were mixed in a 2-1 . round-bottomed, three-necked flask equipped with small Liebig con denser and an electrically heated mantle. The mixture was re fluxed -?3- for one-half hour, allowed to cool, and then shaken with 1 liter of ice water in a 3-1, separatory funnel. The aqueous layer was re moved and discarded; the organic layer was washed with a sodium carbonate solution, dried over sodium sulfate, and distilled through column B to give 1?8 g. (1.^2 moles, ^1$ ) of 2-isopropenylfuran, b.p. 12h,5~12h>r/7hS ram., ng° 1 .1*996. c. In the final step of his procedure, Reichstein*^ hydrogenated 2.5 g« of 2-isopropenylfuran in the presence of a platinum catalyst until the theoretical amount of hydrogen was taken up (one-half hour). Distillation of the product gave an unreported quantity of 2-iso- propylfuran. The present author tried many times to repeat this step, using larger quantities of 2-isopropenylfuran, but without success. Platinum oxide, platinum-on-asbestos, palladiura-on-charcoal and palladiua-on-strontium carbonate were employed as catalysts. Below room temperature (15-20") and at room temperature (25-30 ") no re action appeared to take place. However, if the temperature of the reactants was increased to 1*0 " or above a slow reaction occurred, but no 2-isopropylfuran was formed. A typical hydrogenation procedure is described: A solution of 50«U g. (0*5 mole) of 2-isopropenylfuran in 100 ml. of dry petroleum ether (60-80" ) and 5 g* of palladium-on- strontium carbonate (prepared according to Railings and Smith*®) were placed in an 8-oz. Parr-type bomb wound with fiber-coated 22 E.J. Sailings and J.C. Smith, J. Chea. Soc., 1953. 618. -914- resistance wire for heating purposes (in connection with a variable transformer). The bomb was placed in the rocker carriage, a thermo couple wire was inserted Into the bottle below the surface of the liquid, and the feeder line from the hydrogen source was fitted tightly into place. The bomb was pressured to I4O p.s.i., rocking was commenced, and slow heating was begun. A temperature of $0* was maintained by adjusting the transformer in accordance with frequent readings through a small potentiometer to which the thermocouple was attached. The hydrogen pressure was continually replenished to 50 p.s.i. and was not allowed to fall below 30 p.s.i. "When the theoret ical amount of hydrogen had been absorbed (26 hrs.), the catalyst was filtered, the solvent was removed, and the product was distilled through column A. F orty five grams of material distilled between 120" and 12^". The refractive indices of the samples collected ranged from ng° 1.U3U9 to ng° 1.U379 with no level-off point. A comparison of these properties with the properties of 2-isopropylfuran (b.p. IO6- 109/760 mm., lîp® l.Uh66^®) and 2-isopropenylfuran (b.p. 125-126"/atm., n^® l.i|966®^) indicated that probably the material consisted of a mixture of two or more close-boiling compounds, the most probable being 2-isopropenylfuran, 2-lsopropyltetrahydrofuran (unreported) and 2 -isopropyl-2 ,3-dihydrofuran (unreported). An excellent method of reducing 2-isopropenylfuran to 2-iso- propylfuran was discovered by the present author in the course of the investigation. Sodium and liquid ammonia were found to effect the -95- reduction rapidly, affording a high yield of product. The apparatus setup and experimental procedure for this process has been described in detail on p. 75 in connection irLth the reduction of 2-vinylfuran to 2-ethylfuran. Therefore, only a brief description of a m n is given below; To a stirred solution of g. (3»3 moles) of freshly cut sodium in 1.5 liters of liquid ammonia (contained in a 5-1» flask) was added a solution of 178 g. (1.65 moles) of 2-isopropenylfuran in 1 liter of dry ether. Tfllhen the addition was completed, 18U g. (3»b moles) of powdered ammonium chloride was added and water was introduced into the flask until ammonia no longer refluxed. The aqueous layer was siphoned and discarded and the organic layer was worked up for prod uct. Distillation of this product gave 162 g. (1.5 moles, 905C) of 2-1bo- propylfuran, b.p. 106.8-107.2“/75l mm., n^® 1.U396 (reported^®; b.p. 106-109“/760, mm., ngo I.U466). A total of 280 g. of 2-lsopropylfuran was prepared. 2-tert-Butylfuran.— This furan was prepared in good yield, ac cording to the three-step procedure of Gilman and Calloway^s. A large run which was carried out is described below; a. Since the tert-butylation of methyl 2-furoate was carried out by the same Friedel-Crafts procedure which was used in the attempted isopropylation of this ester (described in detail on p. 8 6 ), only a brief outline of this initial step is given: “•96- A mixture of 63 g* (5 moles) of methyl 2-furoate prepared frran 2-furoic acid (Q.O.and U6S g. moles) of tert-butyl chloride (E.K. "White Label) was added slowly to a stirred, cooled (-5") sus pension of 800 g. (6 moles) of anhydrous aluminum chloride (comm.) in 2.5 liters of carbon disulfide (distilled from phosphorus pentoxide). When the addition was completed the reaction mixture was allowed to stand overnight at room temperature, then was poured over about 1$ kg. of cracked ice and worked up for product* Distillation of this prod uct through column C gave 73$ g* (U moles, 60%) of methyl $-tert- butyl-2-furoate, b*p. 110*/l$ mm#, 1 *1^820 (reported^®: b.p. 110- 11b V 15 mm#, ng® l#b792). b# Seven hundred and thirty-five grams (U moles ) of methyl $-tert-butyl-2*furoate and a solution of 2$6 g. (1 #^ moles) of potas sium hydroxide in lObO g# of absolute ethanol were mixed in a 3-1. single-necked, round-bottomed flask equipped with a small Liebig con denser and an electrically heated mantle# The mixture was refluxed for three hours, cooled, and poured into a 10-1 . crock containing $ liters of ice water. The solution was acidified with cold 6 N. hydro chloric acid and white crystalline $-tert-butyl-2-furolc acid was col lected on a Buchner funnel. It was washed with cold water, pressed as dry as possible, and air dried. The yield of acid was 6?1 g. (b moles, quant. )# Several grams of it was recrystallized from hot water and dried in a vacuum desiccator; m.p. 10$-10$#$" (reported^®; m.p. lOb- 106*). c. A 1-1. three-necked, round-bottomed flask was equipped with an electrically heated mantle, a motor-driven all-glass stirrer, a -97- thermcsneterj and a U"chaped glass tube (2 cm, i.d.) which was led to a 1-1. two-neckcd, round-bottomed flask equipped with a Liebig con denser and cooled by an ice-salt mixture* Two hundred grams of quinoline (E*K. Yellow Label), 20 g. of copper-bronze and 100 g. (0.6 mole) of ^-tert-butyl-2-furolc acid were introduced into the three-necked flaskj slow stirring was commenced, and the mixture was heated. At 210* a smooth, rapid evolution of carbon dioxide took place and the 2-tert-butylfuran distilled into the receiver. When gas evolution had ceased, the flask contents were allowed to cool to 190 ® and 100 g. more of ^-tert-butyl-2-furoie acid was added; the heating process then was repeated. A total of 669 g. (U moles) of acid was decarboxylated in this manner. The receiver contents were distilled through column A to give 392 g. (3.2 moles, 8O/6) of 2-tert- butylfuran, b.p. 117-118®/?^^ mm., ng° 1.^383 (reported^®; b.p. 119- 120 ®/atm., ng® 1 .1*3 8 0 ). A total of 568 g. of 2-tert-butylfuran was prepared. Production of 2-FuryIcycloalkanes 2-FtiryIcyclopropane.— This compound was prepared essentially ac cording to the method used by Kishner®^ for the synthesis of - l-(2-furyl )-2-methylcyc lopropane. 23 H. Klshner, Bull. aoc. chim. France (1*), 767 (1929). -98- a. 3-(2-Furyl)acrolein was prepared according to the procedure of Hinz and co-workers^^, as follows: Into a 12-1. round-bottomed, three-necked flask equipped with a motor-driven Hershberg-type stirrer, a thermometer, and an addition funnel were introduced UOO g. (U*l moles) of freshly distilled fur fural and a solution of 20 g. of sodium hydroxide in 3 *^ liters of water. Rapid stirring was begun and the mixture was cooled to 0** (and maintained at that temperature) by a salt-ice mixture vdiile a solution of 200 g. moles) of acetaldehyde in 1 liter of water was added dropwise through the addition funnel over a four-hour period. After the addition was completed, the solution was neu tralized with glacial acetic acid. The yellow crystals of product were collected on a Buchner funnel, washed with ice water, pressed as dry as possible, and air-dried. Vacuum distillation of this material from a Glaisen flask (into a Dry Ice cooled solids re ceiver) yielded U27 g. (3*5 moles, 8$%) of 3-(2-furyl)acrolein, b.p. 99-101*/lO mm., m.p. aldazine m.p. 166*5 -167 .0 ®; (reported®"^: b.p* 97-102®/lO mm., aldazine m.p. 167-168®). b. A number of attempts were made to prepare 5-(2-furyl)- pyrazoline in a satisfactory yield by Kishner's method®^, but sev eral difficulties were encountered: 1 ) the reaction of 3-(2-furyl)- acrolein with hydrazine always gave a large proportion of the azino, even idien a large excess of the i^drazine was used, 2 ) repeated 24 A. Hinz, G. Meyer and G. Schucking, Bar., 76b. 676 (19b3). -90- vacuum distillations of the impure pyi'azoline resulted in seme de composition each time; a residue of azine always remained, and small amounts of hydrazine were found in the Dry Ice cooled trap. A typical preparation is described below: A solution of .122 g. (1 mole) of 3” ( 2-furyl )ac role in in 150 ml. of absolute ethanol was added rapidly to a swirled, cold solution of 76 g. (2 moles) of anhydrous hydrazine (9^%t Matheson) in 100 ml. of absolute ethanol (contained in a 500-ml. round-bottomed, single necked flask )* The flask then was fitted with a small Liebig con denser and an electrically heated mantle and the mixture was refluxed for three hours. The condenser then was replaced by a modified Glaisen head, the setup was equipped for a vacuum distillation and ethanol, excess hydrazine, and water were distilled from the reaction mixture under reduced pressure (150 mm.). The heavy, yellow oil which remained was transferred to a 250 ml. flask and distilled under reduced pressure through a 1.8 x 25 cm. Vigreaux column to give 80 g. of a pale-yellow oil, b.p. 121-128"/12 mm. A bright-yellow, crystal line residue remained (56 g.) which had a m.p. of 165-168*; it was presumed to be the azine of 3-( 2-furyl )acrolein. A careful, re peated vacuum distillation of the 80 g. of liquid yielded 61 g. of an almost colorless oil, b.p. 121-125"/l2 mm., and left a yellow residue irtiich amounted to 12 g.; the Dry Ice cooled trap contained 3 g. of a liquid which was identified as anhydrous hydrazine. Finally, the 6l g. of material was vacuum distilled slowly at a -100- lower pressure. This time g. of a colorless oil was collected, b.p, U6-ll8"/6 ram., which solidified into a white crystalline solid when cooled below 2^"• Elemental Analysis: C, 60,32%} H, 6 .38^; N, 22,22% (calculated for 5“(2-furyl )pyrazoline, C^HgNgO: C, 61 ,^% } H, S»93%} N, 20.^7%). Therefore, the $-(2-furyl )pyrazoline was still not pure Q f it is assumed to be good material a yield of 0.33 mole or 33% may be calculated, based upon 3-( 2-furyl )acrolei^ ^ The residue firom this final distillation was again a yellow solid (10 g.) and the trap yielded a small amount of hydrazine, indicating that further decom position of the pyrazoline had taken place. c* A platinized-porous-plate catalyst was prepared in Idie fol lowing way: twenty-five grans of small pieces of porous plate were soaked for several hours in a solution of chloroplatinic acid and dried overnight at 110“ in an oven. The yellow-colored fragments then were sealed in a small stainless-steel bomb under an almiosphere of hydrogen and heated at 100* for several hours. By this method a highly active catalyst was obtained. A 250-ml. round-bottomed, three-necked flask was equipped with: 1) a nitrogen inlet tube (capillary) which extended to the bottom of the flask, 2) a small addition funnel, 3) a U-shaped tube (1*5 cm. i.d. ), which was led to a 2$0-ml« round-bottomed, three-necked flask (fitted with a Liebig condenser and cooled by Dry Ice and acetone) -101- h) a rainerai oil bath. Five grains of the catalyst and $ g. of powdered potassium hydroxide were placed in the reaction vessel and heated by the mineral oil bath to 190“. Ten grains of a total of 60 g. (O.Ui mole) of 5"(2-furyl}pyrazoline was added through the addition funnel, a rapid stream of nitrogen was introduced, and the heating was resumed. At a temperature of 200“ rapid decomposition of the pyrazoline took place and the product distilled into the re ceiver; heating was discontinued at this point and the remainder of the furan derivative was added at a rate sufficient to maintain a smooth reaction. Upon the termination of the decomposition, the re ceiver contents were taken up in 100 ml* of ether; the ethereal solu tion was washed with 200 ml. of sulfuric acid, then water, and dried over anhydrous sodium sulfate. An examination of the reaction vessel revealed it to be thinly coated by a black tar. The solutions of product from three runs of the size just des cribed were combined end added through an addition funnel to 300 ml. of rapidly stirred liquid ammonia contained in a 1-1* three-necked, round-bottomed flask equipped for a sodium and liquid ammonia re duction (see p. 70). ïïhen the addition was completed, small pieces of freshly cut sodium were introduced into the mixture until a deep- blue color persisted. Pcwlered ammonium chloride (about 50 g., 1.0 mole ) thai was added from a small Erlenmeyer flask, and water was introduced through the funnel until ammonia no longer refluxed. The -102- . aqueous layer Tas siphoned and discarded, and the organic layer was washed with water and dried over anhydrous sodium sulfate. The ether was removed and the product was distilled slowly through column A to give 37.2 g. (O.3L mole, 2S%) of 2-n-propylfuran, b.p. llii-llli.6"/71j5 mm., np° l.)iiil7 (reported^®: b.p. Uli.5-ll5.5V750,mm., l.h h 5 9 ) and ^$.8 g. (0.52 mole, 395^) of 2-furylcyclopropane, b.p. 327-127»7‘’/ 7b 1 mm., ng° 1.M75. Elemental Analysis; C, 77*17%* H, 7*b8% (calculated for 2-furylcyolo- propane, C^Hq O: C, 77*77%; H, 7.^6%). The purpose of this treatment was to convert the 2-propenylfuran (b.p. 132-133"/7^0 mm.) to 2-n-propylfuran (b.p. 111;.5-115•^‘*/750 mm.) because in an earlier preparation it was found that the furylolefin could not be separated from 2-furylcyclopropane.(b.p* 127-127.7"/7bl mm. ) by distillation through column A due to the closeness of their boiling points (5-6®). However, 2-n-propylfuran has a boiling point 13-lb^ below that of 2-furylcyclopropane and can be separated easily from it. A total of 95 g* of 2-furylcyclopropane was prepared. 2-Furylcyclopentane»— Since the preparation of 2-furylcyclo- pentane had not been reported, it was necessary to find a satisfactory method of obtaining this compounds. Two approaches to its synthesis were considered. 1. The first consisted of a three-step procedure which had been used by Gilman and Calloway^u to prepare 2-isopropylfuran and 2-tcrt- -103- butylfuran. However, all attempts to apply the initial ste p o f that procedure to the preparation of methyl 3-cyclopentyl-2-furoate (by the aluminum chloride catalyzed reaction of methyl 2-furoate and a cyciopentyl halide) were unsuccessful. One of the efforts to effect the Friedel-Crafts reaction is reported here (others were noted in the discussion section): One and one-half liters of purified sym-tetrachloroethane and 175 g. (1.3 moles) of anhydrous aluminum chloride were placed in a 3-1* round-bottomed, three-necked flask equipped with a motor-driven Hershberg-type stirrer, a thermometer, an addition funnel and a gas- outlet tube (loosely plugged with glass wool)* The stirred sus pension was cooled to -5* and maintained at that temperature by a Dry Ice-isopropanol bath while a mixture of 10$ g. (1-mole ) of freshly distilled cyciopentyl chloride and 126 g. (1 mole) of re cently prepared methyl 2-furoate was added slowly through the awidition funnel* Tne cooling bath then was removed and the reaction mixture was allowed to stand overnight at room temperature, after which it was poured (with vigorous stirring) into a crock which con tained 7 kg. of ice* The organic layer was siphoned off, the aqueous layer was extracted with two 200-ml* portions of sym-tetrachloro- ethane, and the combined material (layer and extracts) was washed with 1 liter of 6 H* hydrochloric acid and dried by percolation through a column of anhydrous sodium sulfate. The solvent then was removed and the product was distilled under reduced pressure through -lOli- column B to give g* {0*li2 mole, h2% ) o f recovered methyl 2-furoate, b.p* 82-8j*/20 mm*, l.jjScS (reported^®: b.p. 83- 8L*/20 mm*, l.h36o). A large residue of dark, viscous tar re mained. 2. The second-considered approach proved to be an excellent method of preparing 2-fitrylcyclopentane. The smaller of two runs is described below; a. A solution of n-butyllithium was prepared according to the directions of Gilman and TAright^® : A 3-1. round-bottomed, three- necked flask was equipped with a motor-driven Hershberg-type stirrer, a low-temperature thermometer, an addition funnel, and a Liebig con denser from which a length of tubing was eoctended to the bottom of a nitrogen reservoir ( a 1-gal. bottle). The flask and reservoir were thoroughly flushed with dry nitrogen and 1.^ liters of ether (dried over sodium) was introduced into the flask. During the flushing operation, a coil of lithium wire which weighed about IjO g. was wiped free of grease, washed with dry benzene, then dry ether, and rapidly abraded with a piece of fine sandpaper until all of the oxide coating had been removed. Thirty-four grass (U.9 g. atoms) of this clean wire then was rapidly cut into about 1-cm. pieces vdiich fell directly into the reaction vessel under a slow stream of nitrogen. With stirring, about 10 ml. of a solution of 2lh g. (2 moles) of 25 H. Gilman and G.F. Wright, J. Am. Chea. Soc., Jl, lli99 (19h9)» - 1 0 5 - freshly distilled n-butyl bromide in UOO ml. of sodium-dried ether was added through the addition funnel and the flask contents were cooled by a Dry Ice-isopropanol bath to -10®. In a matter of minutes^ bright spots appeared on the lithium and the solution became faintly cloudy. The remainder of the n-butyl bromide solution then was added over a period of two hours while the Internal temperature was maintained at -10®. After the addition was completed the con tinually stirred reaction mixture was allowed to warm up to 10® over a period of two hours, and the amount of n-butyllithium was deter mined by Gilman and Wright's method®®: the total volume of solution was calculated to be 2.25 1. from graduations marked on the outside of the reaction vessel. A 5"n*l« aliquot sample was withdrawn with a pipet, hydrolyzed and titrated with O.IN hydrochloric acid for total base (using phenophthalein indicator). A second 5-ml. aliquot sample was added rapidly to an ether solution of 3 g* of benzyl bromide, allowed to stand for five minutes, hydrolyzed, and titrated for lithium hydroxide. The difference in the two titrations gave the amount of organometalllc present (O.OOUoU moles/5 ml. or 1.32 moles/ 2.25 litersj 90%, based upon n-butyl bromide). b. To the above-prepared solution, I36 g. (2 moles) of freshly distilled furan was added at a fairly rapid rate, with stirring, through the addition funnel. A mild exothermic reaction caused the ether and/or furan to reflux gently; no cooling was necessary. The solution slowly changed from a light-blue to a light-orange color - 106- in about one-half hour after the addition was completed. In their preparation of 2-furyllithiiim, Benkeser ond Currie^® had added a n-butyllithium solution to furan, then refluxed the mixture for three hours and allowed it to stand overnight at room temperature. There fore, the presently described reaction mixture was treated in the same manner (heat being applied by an electrically heated mantle). The excess lithium metal from the n-butyllithium preparation then was removed in the following manner: the stirrer was withdrawn and the surface of the solution was skimmed with a flat scoop made from a short length of one-quarter inch dowel and a semi-circle of thin glazed cardboard (of one-half inch radius). The yield of 2-furyllithium was determined as 1.1:5 moles j 80/6, based upon n-butyllithium by the titration method which was used to determine the yield of n-butyllithium. c. The 2-furylllthium solution was cooled initially to 10" by a cold water bath, stirring was commenced, and a solution of 122 g. (1..U5 moles) of freshly distilled cyclopentanone in 200 ml. of dry ether was added at a moderate rate through the addition funnel. By the time the addition was completed (one-half hour) the vessel con tents had warmed to 35** The bath was removed, an electrically heated mantle was fitted to the flask, and the reaction mixture was refluxed for two hours. It then was allowed "to cool and a sufficient amount of 26 R.A. Benkeser and R.B. Currie, J. Am. Chem. Soc., JO, 1?80 (19U8). -107- water was added through the addition funnel to cause the lithium hydroxide to separate as small pellets. 'ITie ether solution was de canted, the pellets were washed with two lOOml. portions of ether, and the combined material (extracts and solution) was dried over anhydrous sodium sulfate. The solvent was removed and the product was distilled slowly under reduced pressure through column 3* A considerable amount of dehydration of product occurred during the distillation, which is outlined here; Fractions 1-7 distilled between and 52" at 2 mm. pressure and exhibited refractive indices which ranged from 1.5398 to 1.5U03 each fraction first was dried over anhydrous sodium sulfate before its refractive index was determined. These fractions were combined as fairly good l-(2-fuiyl)cyclopentene (108 g.^ 0.3 mole, 55^ crude) and added later to material obtained by dehydrating l-(2-furyl)cyclopentanol (see step d.). Fraction ^ an intermediate fraction, distilled from U8" to 73" at 2 mm. pressure. Weight: 5 g. Fractions 2 and 10, proved to be good l-( 2-furyl )cyclopentanolj they distilled at a constant 73"/2 mm, and had the same refractive index, n^° 1.5072, Combined weight: 27 g* (0.18 mole, 12%)» Elemental Analysis ; C, 71,12^j H, 7.97^ (calculated for l-( 2-furyl )- cyclopentanol, CgHigOg: 0, 71.02^; H, 7.95^). A residue of only 2 g. was obtained and the Dry Ice cooled trap yielded no additional organic material. -lo8- d. Tvrenty-seven gï*am3 (0.18 mole) of 1-(2-furyl)cyclopentanol, 13.5 G» (0«18 mole) of acetic anhydride (commercial) and g. (0.09 mole) of anhydrous sodium acetate were mixed in a ICO-ml. round- bottomed, single-necked flask equipped with a Liebig condenser and an electrically heated mantis. Tlie mixture was refluxed for one- half hour, cooled, and poured into a separatory funnel v/hich con tained 100 ml. of water. One hundred millilitei's of ether was added, the mixture was shaken, the aqueous layer was withdrawn and discarded and the organic layer was washed vdth a solution of sodium bicarbonate and dried over anhydrous sodium sulfate. The solvent was removed and the product was addod to the 108 g. of furylolefin Tdiich had been obtained as fractions 1-7 in the l-(2-furyl)cyclopentanol distillation. A careful distillation of this combined material under reduced pressure through column B gave 106 g. (0.78 mole; based upon 2-furyllithium) of l-(2-furyl)cyclo- pentene, b.p. l4Ü°/2 mm., n^° 1*5398. Elemental Analysis: C, 80.cO; H, 7*61^ (calculated for 1-(2-furyl)- cyclopentene, Gg%oO: C, 80.6{^; H, 7»Ü6^). The dehydration procedure just described had been used earlier by Bachman and Heisey to dehydrate dimethyl-2-furylcarbinol. 3 . Into a 5-1. round-bottomed, three-necked flask equipped for a sodium and liquid ammonia reduction (see p. 75) was introduced 1.5 liters of liquid ammonia and bo.l g. (2.82 moles) of freshly cut sodium. Stirring was begun sid, after allowing the sodium to dissolve - 1 0 9 ” completely, a solution of 23Ü g. (1.7L moles) of li*(2-furyl)cyclO“ pentene in 1 liter of ether was added slowly through the addition runnel. Then the addition was completed and the mixture had stirred for 15 min. longer, the addition funnel was removed from the setup and 216 g, (L moles) of powdered ammonium chloride was added slowly from an Erlenmeyer flask (the neck of which was bent 90® from the normal) attached to the reaction flask with a section of Gooch tubing. The addition funnel was replaced and water was added through it until ammonia no longer refluxed. Stirring was discontinued, the aqueous layer was siphoned and discarded and the organic ]ayer was washed with water and dried over anhydrous sodium sulfate. The sol vent was removed and the product was distilled under reduced pres sure through column B to give 22h g. (1.6ii moles, 95%) of 2-fury 1- cyclopentane, b.p. 96.U®/65 mm., n^° l.h8i^9# Elemental Analysis ; C, H, 8.81$ (calculated for 2-furylcyclo- pentane, CgHiaO: C, 79»29%$ H, 8.88$). A total of 22lj. g. of 2-furylcyclopentane was prepared. 2-Furylcyclohexane. — As in the case of the 2-furylcyclopentane preparation, initial efforts were made by the present author to syn thesize 2-furylcyclohexane according to the procedure which had been used by Gilman and Calloway^® to prepare 2-isopropylfuran and 2-tert- butylfuran. However, all attempts to apply the initial step of that procedure to the preparation of methyl 5“cyclohexyl-2-furoate (by the aluminum chloride catalyzed reaction of methyl 2-furoate with a -110- cyclohexyl halide) proved unsuccessful; only polymer and unreacted methyl 2-furoate were obtained. Since the results of this part of the investigation closely paralleled those which were encountered in the attempts to prepare methyl 5”Cyclopentyl-2-furoate, no example will be given here, but reference is made to p. 103 and the dis cussion section, pp. 2^-27* Tlie preparation of 2-furylcyclohexane, a new 2-substituted furan, was successfully carried out by the same steps which were used in the preparation of 2-furylcyclopentane (described in detail on pp. lOU to 10Ç). Therefore, the single run which was made is presented briefly below; a. A solution of n-butyllithium was prepared as follows: to a stirred suspension of 19 g. (7 g* atoms) of 1-cm. lengths of clean lithium wire in 2*i? 1. of sodium-dried ether (contained in a ^-1. round-bottomed, three-necked flask) was added 30 ml. of a solution of 111 g. (3 moles) of freshly distilled n-butyl bromide in $00 ml. of sodium-dried ether. The vessel contents then were cooled to -10® and when the reaction began, the remainder of the halide solution was added over a three-hour period. Cooling was discontinued, the inter-- nai tanoerature was allowed to rise to 10®, and a determination of the yield of n-butyllithium was made (2.6 moles; 86.^%, based upon n-butyl bromide )• b. Two hundred and four grams (3 moles) of purified furan was added to the stirred n-butyllithium solution without cooling. When -111- tbe addition was completed (about one-half hour ), the mixture was stirred for one-half hour, refluxed for three hours, and then allowed to stand overnight. The excess Uthium was removed and the yield of 2-furyllithium was determined (2 moles; 1$%, based upon n-butyllith ium ). c. The 2-fuiyllithium solution was cooled Initially to 10®, stirring was commenced, and a solution of 1$6 g, (2 moles) of puri fied cyclohexanone in 200 ml. of dry ether was added at a moderate rate. Upon completion of the addition the temperature of the re action mixture had risen to hO®. This mixture was refluxed for two hours, allowed to cool to room temperature, and water was added until the lithium hydroxide settled in the form of small pellets. The ether solution was decanted, the pellets were washed with 2 200-ml. portions of ether, and the ccaubined material (extracts and solution) was dried over anhydrous sodium sulfate. The solvent was renoved and the product was distilled under reduced pressure through column B* This distillation is outlined below; Fractions 1-9 distilled between 91.1* and 97.6* at 3 mm. pres sure. They exhibited refractive indices ranging from 1.^122 to 1.^217 (n^° ); each fraction was dried over anhydrous sodium sulfate before its refractive index was determined. Combined weight: 32 g. Fraction 10 distilled over a range of 1,9* (97.6-99.5*); it partially solidified in the bottle in urtiich it was collected. Weight: 66 g. - U 2 - Fractions 11-13. good l-(2-furyl)cyclohexHnol, distilled between 9SP,5* and 100.1* • Each solidified in the bottle in which it was col lected (m.p. 36-37*)• Combined weight; 119 g* (O.7I mole; 3à%, based upon 2-furylllthium)* Elemental Analysis; C, 72.31%; H, 8.53% (calculated for l-(2-furyl)- cyclchexanol, C10H14O2 * C, 72.26%; H, 8.1|9%)* A residue of only 3 £• was obtained and the Dry Ice cooled trap yielded no additional organic material* d. All of the fractions iriiich were collected in the above distil lation were combined as pure l-( 2-furyl )oyclohexanol for the dehydra tion step* Two hundred and sixty-seven grams (1*6 moles) of this al cohol, 163 g. (1*6 moles) of acetic anhydride (Carbide and Carbon), and 65 g* (0*8 moles) of anhydrous sodium acetate were mixed in a 1-1. round bottomed, three-necked flask. The mixture was refluxed for one-half hour, cooled, and poured into a separatory funnel which contained 500 ml. of water. Two hundred milliliters of ether was added, the mixture was shaken vigorously, and the aqueous layer was withdrawn and discarded* The ether solution then was washed with 500 ml. of a 5% solution of sodium bicarbonate and dried over an hydrous sodium sulfate. The solvent was removed and the product was distilled under reduced pressure through column B to give IU7 g, (1 mole; 50%, based upon 2-furylllthium) of l-(2-furyl)cyclohexene, b.p. 8O.2 V 3 nra., n§° 1 *5U1 5 . - u > Elemental ^alysls; C, 8l.28^j H, Q,CCt^ (calculated for l-(2-furyl)- cyclohexene, CiçHxaO; C, Bl.OL^j H, 8.17$)« e. To a stirred solution of $0,6 g. (2.2 moles) of freshly cut sodium in 2 1. of liquid ammonia (contained in a p-l. round-bottomed, three-necked flask) was added slowly a solution of H j7 g. (l mole) of l-(2-furyl )cyclohexene in $00 ml. of dry ether. Stirring was main tained for one-half hour longer, then 111 g. (2.2 soles) of powdered ammonium chloride was introduced. Water was added until ammonia no longer refluxed, the aqueous layer was siphoned and discarded, and the ether solution was washed with water and dried over anhydrous sodium sulfate. .The solvent was ranoved and the product was distilled under reduced pressure through column B to give 1L6 g. (0.97 mole, 97$) of 2-furylcyclohexane, b.p. 109®/$0 mm., 1 Jj873 . Elemental Analysis: C, 80.03$j H, 9«1:5$ (calculated for 2-furyIcyclo- hexane, CioHi^O: C, 79.96$; H, 9.39$). A total of 1U6 g. of 2-furylcyclohexane was prepared. Production of Miscellaneous 2-Substituted Fur ana Furfurylidene Pi acetate^!— In a 300-tsl. Glaisen flask were mixed 102 g. (1 mole) of acetic anhydride (Carbide and Carbon) and 0.1 ml* of concentrated sulfuric acid. The mixture was cooled to 10* by an ice bath and 96 g. (1 mole) of freshly distilled furfural was added 27 R.T. Bertz, Organic Syntheses. Vol. 33, John Wiley and Sons, Inc., New York, N.T., 1953, p. 39. -lüj- over a ten-minute period while the temperature was maintained at 10» 15*. After the addition was completed, the flask contents were mixed well by swirling, the cooling bath was removed, and the reaction mix ture had cooled to room temperature, 0«L g« of anhydrous sodium ace tate was added, the flask was fitted for a vacuum distillation (in cluding a mineral oil bath), and 2-furfurylidene diacetate was distilled from the vessel. One hundred and forty grams (0.7 mole, 70?) were obtained; b.p* lb0*/20 mm., m.p. 52-52.8* (reported®*^: b.p. lii0-lU2“/20 ran., m.p* 52-53*)» A total of lUO g. of furfuiylldene diacetate was prepared. Furfuryl Acetate^^, — A mixture of 500 ml. of dry benzene, 300 g. (3 moles) of freshly distilled furfuryl alcohol, 113 g* (l*li moles) of anhydrous sodium acetate and 325 g» (3*2 moles) of acetic anhydride (Carbide and Carbon) was placed in a 3-1» round-bottomed, three-necked flask fitted with a rubber stopper, a motor-driven Hershberg-type stirrer and a small Liebig condenser (provided with a calcium chloride tube). The flask was heated on a steam bath for three hours, with stirring. After it had cooled to roam temperature, the reaction mix ture was poured into a i^-l. separatory funnel and shaken vigorously with 2 liters of cold water. The aqueous layer was withdrawn and dicarded and the organic layer was washed with 500 ml. of a 5? sodium carbonate solution, then with 2 liters of water. The benzene solvent was removed and the product was distilled under reduced 28 A.A. Sunier, Organic Syntheses. Coll. Vol. I, John Wiley and Sons, Inc., New York, N.Y., 1932, p. 279. - 1 1 5 - pressure through column B to give 336 g. (2.h moles, 80%} of furfuryl acetate, b.p. 69-70*/7 mm. (reported*®: b.p. 69-70/7 mm.). A total of 336 g. of furfuryl acetate was prepared. 2-Bromofuran.— a. The brcaaination of 2-furoic acid, when carried out according to the pix>oedure of Moldenhauer and co-workeira®®, in variably gave a low yield of 5"bromo-2-furoic acid. Consequently, their process was modified by the present author to give good yields of product. This modified method was used in the run described i below: One hundred and twelve grams (1 mole) of 2-furoic acid (Q.C. ) was dissolved (warming necessary) in 360 ml. of glacial acetic acid (du Pont C.P. ) and the solution was poured into a 1-1. three-necked, round-bottomed flask equipped with a separatory funnel, a motor- driven Hershberg-type stirrer, and a large Liebig condenser (from irtiJcfa was led a tube to a gas-absorption trap filled with aqueous sodium hydroxide). Stirring was commenced and 208 g. (1.3 moles) of bromine was added dropwise through the funnel at a moderate ratej no cooling was found to be necessary since the temperature of the reaction mix ture never exceeded 5^** After tdie addition was completed the warm mixture was refluxed for one-half hour, cooled to 100", and poured into 2.5 liters of boiling water in irfiich 20-30 g. of activated charcoal was suspended. A jet of steam was passed into the mixture for ten minutes in order to agitate it and also maintain the 100" *9 0 . Moldenhauer, et ^ . , Ann., 560. 188 (1953)* - U 6 - temperature. The solution then was filtered from the charcoal through a Bucimer funnel and cooled bj' ice water to give 133 g* (0.7 mole, ^0%) of white, crystalline $"bromo-'2-furoic acid, m.p. 188-169* (reported^®; m.p. 18?*)• b. 5“Bromo-2-furoic acid was decarboxylated by the method which was used to decarboxylate ^-tert-butyl-2-furoic acid (see p. 97)* In a typical run 287 g. (1*5 moles) of 9“bromo-2-furoic acid was decarboxylated in hO~SO g. portions at 210* by ID g. of copper-bronze in 100 ml. of quinoline (E.K. Yellow Label). Distillation of the product through column A gave 110 g. (O.?^ mole, $0%) of 2-broœofuran, b.p. 101-101.2*/7U9 mm., n|° l.US>82 (reporteds®: b.p. 101.9-102.2*/ 7U* mm., n^° 1.1(9809 )• Upon standing for several days at room temper ature in a dark bottle but with no polymerization inhibitor, about one-fourth of the 2-bromofuran polymerized. Even when stored in a refrigerator in the presence of an inhibitor, a slow darkening took place. A total of 350 g. of 2-bromofuran was prepared. Tetramethylfuran.— a. 3,U-Dimethyl-2,9-hexanedione was prepared by the method of Moore^^, as follows: A mixture of 3^0 g. (5 moles, a large excess) of freshly dis tilled methyl ethyl ketone and 183 g. (1.29 moles) of dl-tert-butrl peroxide (Shell) was placed in a 1-1. stainless-steel bomb, which was equipped with a needle valve, evacuated, and fitted into a 3 0 A.F. Shepard, N.R. Winslow, and J.R, Johnson, J. Am. Chem. Soc., 2083 (1930 ). 3 1 C.Q. Moore, J. Chem. Soc., 1991. 236. -117- rocker carriage (irtiich contained its own heating jacket). The bomb then was heated at liiO* for 2k hrs*, without rocking; the temperature was controlled by means of a thermocouple and an electrical regulator. At the end of the reaction period the bomb was cooled by running tap water and the valve was opened to the atmosphere. The reaction mix ture from this run was combined with those of three other runs and the combined material was distilled under reduced pressure through column C to give 191 g* (1*3); moles; 2^/6, based upon dl-tert-butyl peroxide) of 3,ii-dimethyl-2,^-hexanedione, b.p. ?8-80"/8 mm., n^° l.L3&2-l.ü3Wi (reported^^: b.p. 7b-76*/8 mm., ng° l.L3i*2). The second step of this procedure was carried out according to Oaertner and Tonkyn®*; b* One hundred and ninety-one grams (1.3U moles) of 3>U"dimethyl- 2,S-hexenedione and 180 g. (1.7 moles) of acetic anhydride (Carbide and Carbon) were mixed in a ^00-ml. single-necked flask equipped with a large Liebig condenser and an electrically heated mantle. Through the top of the condenser, was introduced ^ g. of anhydrous zinc chloride all at once. In about five minutes a very exothermic re action ensued which caused the mixture to reflux vigorously. After the initial reaction had subsided, the mixture was refluxed for two hours, allowed to cool, and poured Into a 2-1. separatory funnel which contained a liter of cold water and 900 ml, of ether. The funnel was shaken vigorously, the aqueous layer was removed and dis carded, and the organic layer was washed with 200 ml. of a 5/É sodium 32 R. Oaertner and R.G. Tonkyu, J. Am. Chem. Soc., 21» 9872 (1991). -118- carbonate solution and dried over anhydrous sodium sulfate. The ether solvent was removed and the product was distilled under reduced pressure through column B to give 7Ü.5 g. (0.6 mole, k5%) o f tetra methylfuran, b.p. 62*/30 mm., ng® lJi572 (reported^*: b.p. lU^-liib*/ 7hd mm., ng° l.US^O). A total of 7I4.5 8* of tetramethylfuran was prepared. Production of Diels-Alder Adduets of 2-AlkyIfurans and 2-Kethyl- ^-alkylfurans with Malelc Anhydride In that part of the discussion section of this dissertation which dealt with the preparation of the Diels-Alder adduets of 2-alky Ifur ans and 2-methyl-^-alkylfurans with malelc anhydride (pp. 32 to 3 6 ), it was explained that several modifications of the usual procedures for preparing these compounds were made. The first modification was em ployed in order to insure the formation of finely divided adduets in those cases where hard, not easily pulverized crystalline material would otherwise be formed. It was explained later (p. I4.3 ) that this finely divided state of the adduets was necessary in order for their aromatization to the corresponding 3-alkyIphthalic anhydrides and 3-methyl-6-alkylphthalic anhydrides to be carried out smoothly. Other modifications were employed in order to expedite the formation of Diels-Alder adduets, ^ Consequently, each 2-alkylfuran and 2-methyl-5-alkylfuran was placed in one of four classes (see Table IV, p, 3$), based upon its -119- rate of reaction Trlth malelc anJiydride and the physical condition of the product. A typical preparation of the adduct of a furan compound in each of the classes is given below (exception: the preparations of the adduets of two furans in class 1 are given). All Diels-Alder reactions were carried out in an efficient hood. Total quantity of adduct prepared: Class 1 2-Metbylfuran $280 g. 2-Ethylfuran 1826 g. 2,S-Dimethylfuran 6o80g. 2-Hethy1-5-0thyIfuran 950 g. Tetramethylfuran 91 g* 1* Four hundred and fifty grams (5*1 moles) of purified 2-methyl- furan, U90 g. (5 moles) of powdered, recrystallized aaleic anhydride, 0*5 g* of hydroquinone, and 2 liters of ether ware miaæd in a $ -1 , stainless-steel beaker, and then placed in a warm water bath (60* ). The mixture was hand-stirred with a glass rod until all of the malelc anhydride had dissolved and a mild eju*thermic reaction had begun (about 10 minutes)* The beaker was removed from the bath, a ther mometer was inseirted into the solution and the mixture was allowed to stand until the reaction had ceased (one-half hour, as evidenced by a drop in temperature ). Ether was replenished as it was lost during the course of the reaction. When cooling had conmienced, the vessel was clamped to a ring stand and a motor-driven Hershberg-type stirrer -120- iras inserted into the solution* At the first sign of crystalliza tion of product, rapid stirring was begun and maintained for about hS minutes (the beaker was cooled with ice water during the last 20 minutes). The finely granular adduct was collected on a Buchner funnel, washed with 200 ml. of ice-cold ether, and air dried. Eight hundred and forty-six grams moles; 90jt, based upon maleic an hydride) of white 3-methyl-3, b-epoxyy^-tetrahydrophthalic anhydride were obtained. The filtrate was concentrated to one-fourth volume and cooled by an ice-salt mixture to give 75 g* (0*U mole, 8^) of pale-yellow crystals of product* 2* Sixty-two grams (0*5 mole) of freshly prepared tetramethyl- furan, 0,1 g. of hydroquinone, and 300 ml. of ether were mixed in a 1-1. stainless-steel beaker; a thermometer was inserted and the beaker was placed in an ice water bath. To the cooled, hand-stirred solu tion was added 1|6 g. (0.U8 mole) of powdered maleic anhydride (re- cryst.). As the anhydride dissolved, a mild exothermic reaction ensued; it continued for five minutes after dissolution was complete. Wo appreciable amount of ether was lost. TShen a thermometer reading indicated a rapid drop in temperature, the beaker was removed from the bath and clamped to a ring stand. The remainder of the procedure was carried out in the manner described on p, 115 (2-methylfuran- maleic anhydride adduct preparation). The finely granular adduct was collected on a Buchner funnel, washed with 50 ml. of ice-cold etlior, and air dried. Ninety-one grams (O.Ul mole; 85%, based upon maleic -121- anhydride) of white 3,14,5 ,6-te trame thyl-3, 6-epoxy-^-te trahydro- phthalic anhydride was obtained. The filtrate was concentrated to one-fourth volume and cooled by a s<-ice mixture to yield 10 g. (O.OU5 mole, 9%) of pale-yellow crystals of product. Total quantity of adduct prepared: Class 2 2-n-Propylfuran I66I4 g. 2-Isopropylfuran 396 g. 2-Furylcyclopropane 92 g* 2-Furylcyclopentane 113 g. 2-Furylcyclohexane III4 g« Three hundred and sixty-three grams (3«3 moles) of freshly pre pared 2-n-propylfuran, 3li| g. (3.2 moles) of powdered maleic anhydride (recryst.), 0*1 g. of hydroquinone, and 1.5 liters of ether were mixed in a 3-1. Pyrex-glass beaker, which was placed in a warm water bath )• The mixture was hand-stirred with a glass rod until all of the anhydride had gone into solution and a mild exothermic reaction had begun (immediately afterward fine white needles rapidly began to settle from solution). The beaker was removed from the bath, a thermtmieter was inserted into the mixture, and it was allowed to stand until a fall in temperature indicated the reaction to be complete (about one-half hour, during which time about 150 ml. of ether was replenished). The vessel was placed in a refrigerator for three hours, then the product was collected on a Buchner funnel, washed -122- with 200 ml. of ice-cold ether, and air dried. Six hundred and three grains (2.9 moles; ?2%, based upon maleic anhydride) of white 3-n-propyl-3,6-epoxy-^‘*-tetrahydrophthalic anhydride was obtained. The filtrate was concentrated to one-fourth volume and cooled by an ice-salt mixture to give 30 g. (O.lL mole; S%) of pale-yellow needles of product. Total quantity of adduct prepared: Class ^ 2-Bromofuran 237 g* Furfurylidene diacetate 137 g* Furfuzyl acetate 2iiO g* One hundred and forty-seven grams (1 mole) of 2-bromofuran, 96 g• (0.98 mole) of powdered maleic anhydride (recryst* ), a trace of hydro quinone, and k$0 ml. of ether were mixed in a 1-1. I^ex-glass beaker, which was placed in a warm water bath (60®)» The mixture was hand- stirred with a glass rod until the maleic anhydride had dissolved. The reaction vessel was allowed to stand in the bath for one hour longer (dinring which time about 100 ml. of ether had to be added), then for several hours at room temperature, and finally overnight in a refrigerator. % e product was collected on a Buchner funnel, washed with IDG al. of ice-cold ether and air dried. Two hundred and thirty- seven g r a æ (0.96 mole; 97%* based upon maleic anhydride) of 3^bromo- 3,6-epoxy-^^tetrafaydrophthalic anhydride was obtained as fine, white needles. The filtrate was concentrated to one-fourth volume and cooled by an ice-salt mixture to give 3 g. of additional material. -123- Total quantity of adduct prepared: Class 4 2-Methyl-5-n-propylfuran 990 g. Three hundred and seventy-two grams (3 moles) of 2-methyl-$-n-pro- pylfuran, 285 g. (2.9 moles) of powdered maleic anhydride (recryst.), 0.1 g. of hydroquinone and 1*5 liters of benzene were mixed in a 3-1» round-bottomed, three-necked flask equipped with a small Liebig con denser and an electrically heated mantle. The mixture was refluxed for three hours and then poured into a 2-1. Pyr ex-glass beaker. The beaker was placed in a refrigerator for two hours, during which time fine white needles of product settled out. The material was collected on a Buchner funnel, washed with 200 ml. of ice-cold ether, and air dried. Five hundred and ten grams (2*3 molesj 80$, based upon maleic anhydride) of 3-methyl-6-n-propyl-3,6-epox7^/5^-tetrahydrophthalic an hydride was obtained. The filtrate was concentrated to about one- fourth volume and cooled by ice water to give an additional 102 g. (0.U6 mole, 16%) of product (pale-yellow, fins needles). "flie yields of adduct s of the fur ans in all four classes were very good; they ranged from 92$ to essentially quantitative. 2-Tert-butvlfuran would not react with maleic anhydride. At tempted preparations of the adduct are listed below. 1. The furaa, maleic anhydride, trace of hydroquinone were warmed in ether. 2. The fur an, maleic anhydride, trace of hydroquinone were heated in benzene at reflux. -12l- 3. The f11 ran, maleic anhydride, trace of hydroquinone were heated in toluene at reflux, if. The furan, maleic anhydride, trace of hydroquinone, benzene solvent were heated at 150“ in a bomb under a nitrogen atmosphere. 2-Tert-butylfuran also would not react with maleimide when a solution of the reactants in benzene was refluxed. Elemental analyses (carbon and hydrogen) of seven new Diels- Alder adducts have been obtained. These analyses are given in Table IX. The melting points of the adducts which were prepared during the investigation are presented in Table X* Neutral equivalents of adducts were determined as a measure of purity of product after each preparation. In addition, neutral equiv alents were determined on recrystallized samples to be used for melting point and elemental analysis determinations (the values ob tained were within 1-2% of the calculated values in all cases). It is not considered necessary to record these values here* Table IX Elemental Analyses of New Diels-Alder Adducts Furan-maleic Anhydride Analyses %eoretloal ______Adduct______i c % Ü______|__c______$_H______ 2-n-Propylfuran 63.65 6.01 63 .L5 5*81 2-Isopropylfuran 62.96 5,81 63 .h5 5*81 -125- Table IX (Continued) Furan-iîaleic Anhydride Analyses Theoretical Adduct % C % U % C % W 2-Fuiylcyclopropane 63.93 5.01 61.06 Ü .90 2-FuryIcyclopentane 66.62 5.95 66.65 6.02 2-Fury Ic yc lohexan e 67.68 6 .U5 67.72 6.50 2-Methy1-5-sthylfuran 63.L5 5.77 6 3 .1*5 5.81 2-Me thyl-5-n-propylf uran 6U .85 6.26 61*.35 6.35 Table X Melting Points of Diels-Alder Adducts Furan-Maleic Anhydride Melting Pointé’ Melting Point Adduct (Determined) (Reported) 2-Methylfuran 82-33" 80" (ref. 33) 2-Ethylfuran 97.2-98" 97-93" (ref. 31*) 2-n-propylfur an 79 .1-80" 2-lBopropylfuran 89-90 " 2,5-DimetoyIfur an 7 6 .5 -77“ 78" (ref. 33) 2-Msthyl-5-ethylfuran 8 1 .5-33 " 2-Hethyl-5-n-propylfuran 6 0 .5 -6 1 .5 " 2-Fury Icyclopropane 81-32" of chloroform (or ether) at roan temperature, adding petroleum ether (2-3 Tola. ), and cooling the solation slowly. 33 K. Alder and K. Backondorf, Ann., 535, 101 (1938). 34 R. Paul, Bull. 800. chia, France, 10. I63 (19li3). -126- Table I (Continued) Furan-Maleic Anhydride Melting Point ^ Melting Point Adduct (Determined) (Reported ) 2-Furylcyclopentane 90-90.5" 2-Furylcyclohexane 9Ü-9U.5" 2-Bromofuran 111.5-115.5" 116" (ref. 35) Furylfuryl acetate 115-115.5" 111" (ref. 3 6 ) Furfurylidene diacetate 125-126" 126.5-127" (ref. 35) Tetramethylfuran 96-97* 95.3-96.6" (ref, 32) ^ Samples were reciystallized by dissolving them in a minimum amount of chloroform (or ether) at room temperature, adding petroleum ether (2-3 vols.), and cooling the solution slowly. Production of 3-Alkylphthallc Anhydrides and 3-îfothyl-6-alkylphthallc Anhydrides Hydrogen Bromide Method.— A number of unsuccessful efforts were made to effect the aromatization of several Diels-Alder adducts by reacting solutions of them with hydrobromic acid (L8%) or anhydrous hydrogen bromide in acetic acid. These attempts have been recorded in Table 7 (p. Ul) in the discussion section. One experiment is des cribed below: Eighteen grams of hydrobronic acid (0.1 mole HBr) was added to a solution of eighteen grams (0,1 mole) of 3-z2thyl-3,6-epoxy^- 36 M.G* Van Caaqpen and J.R, Johnson, J. Am. Chem. Soc., L30 (1933). 36 0. Diels, et Ann., U90, 2h3 (1931). -127- tetrahydrophthalic anhydride In 2^0 ml, of glacial acetic acid (con tained in a ^00-ml. Erlenmeyar flask ) and the solution v?as heated for 2 hrs. on a steam bath. During this period the color of the solution changed from black to dark brown and a light-browm finely divided substance settled. The mixture was cooled and the brovm solid was collected on a Buchner funnel, washed with cold water and dried. It was not soluble in hot benzene, absolute ethanol, or petroleum ether (60-80 " ), but was soluble in hot water. The aqueous solution was de colorized with charcoal and cooled to give 1|.2 g. of a white powder (which later was identified as fumarlc acid through a comparison of its infra-red spectrum with that of an authentic sample ). The filtrate of the reaction mixture then was poured into $00 ml, of ice water; a brown polymer precipitated. It was collected on a Buchner funnel, washed with cold water, and dried in a vacuum desiccator. The substance dissolved readily in benzene to give an almost black solution. The solution was cooled at 10" for 2h hrs., but only a slight amount of black tar settled. Sulfuric Acid Method.— The successful application of this method to the aroma tization of a number of adducts has been described in detail in the discussion section of this dissertation (pp. h3 to 5 1 ). All experiments which were conducted, including conditions and yields of products, have been recorded in Tables VI and YII (pp. to L9). One of these experiments is given below; -128- Five and one-half liters of concentrated sulfuric acid (!?6 ,6$^) was placed in a 12-1 * round-bottomed, three-necked flask equipped rrlth a motor-driven glass rod ( to vdiich was attached a large Teflon blade), a low-temperature thermometer, end a large povfdor funnel. The stirred acid was cooled to -10® by a Dry Ice-icopropanol ba+h and H 6ii g. (6 moles) of finely granular 3 ,6-dijnethy1 -3 ,6-epoxy- a*" tetrahydrophthalic anhydride was added, with rapid stirring, over a period of two hours. The temperature of the solution was !cept be tween -10® and -1?® throughout the addition and for one hour after ward. The bright-red solution then was allowed to warm to 5° (with out removing the cooling bath) and equal portions were poured into two 5 "gal. crocks filled with ice, with vigorous stirring (a glass rod of one-inch diameter was found to be useful). The two mixtures were stirred occasionally until most of the ice had melted. The light tan product was collected on a Buchner funnel, washed thoroughly with ice water, and pressed as dry as possible. The material then was stirred in a cold 10$ solution of sodium bicarbonate until no further evolution of carbon dioxide could be detected. The residue was collected on a Buchner funnel, washed with cold water, pressed as dry as possible, and air dried* It then was recrystallized from a benzene-petroleum ether (60-8O®) mixture (after decolonization with charcoal) to give 5U6 g. (3,1 moles, ol$) of white, crystalline 3 ,6-dimethylphthalic anhydride. Several grams were recrystallized from benzene-petroleum ether (60-80 ®) for characterization purposes: -129- m.p. 1Ü2-1L2.3" (reported^^: m.p. 1142-1113“ )o The sodium bi carbonate solution was neutralized to give h$ g. (0.3 mole, of 2,5-dimethylbenzolc acid^ m.p. 131.9-132“ (reported^'': m.p. 132-1314"). In the discussion section (p. $0) it was mentioned that in addition to the anhydrides, the 2-methyIfuran-, 2-ethylfuran-, and 2-n-propylfuran-maleic anhydride adducts also yielded small amounts of the m-alkylbenzoic acids when they were aromatized in sulfuric acid. Since the arcmatization procedures are not to be described, mention is made here of these results and the yields and physical properties of the acids are given (Table XI). Also included in the table is 2,^-dimethylbenzoic acid* In each case, the yield of acid was actually determined in only two experiments. Table XI Furan-maleic Melting 4(" Melting Anhydride Acid Point Point Yield A d d w t (Found) (Reported) 2-Methylfuran m-Toluic II2 .9 -IIÜ" im.i4-ll9"(ref.38) 9$, 6% 2-Ethylfuran m-Ethyl- 148.9 -148.8 " U7-U7.6"(ref.39) $% benzoic 2-n-Propyl- m-n-Propyl- 143.2-143.9“ I43" (ref.l40) h%. 7% furan benzoic 2,9 -Dimethyl- 2,9-Dimetbyl- 1 31 .9 -132“ 132-1314“ (ref .37) 6% furan benzoic ^ Melting points are of recrystallized samples of the first yield recorded. 37 M.S. Newman and B.T. Lord, J. Am. Chem. Soc., 73U (19ijij ). as H.A. Smith and J.A. Stanfield, J. Am. C h m . Soc., jl, dl (19U9)i as M.S. Newman, J. Am. Chem. Soc., 376I4 (1999)# 40 M. Crawford and F.H.C. Stewart, J. Chen. Soc., 1992. 286. -130- The sodium bicarbonate washings of all other aromatized adducts yielded only polymers when acidified. In Table VII (p. h8) was recorded the anomalous behavior exhib ited by 3-isopropyl-3;6-epoxy-^'*-tetrahydrophthalic anhydride when it was aromatized. It should be mentioned here that the workup proce dure used in this case was identical to the one used in all others* T?hen the solid material from the aromatization (which included much polymer) was treated with a cold sodium bicarbonate solution, it re acted completely. Careful addition of acetic acid precipitated trtiite crystals of 3'*isopropylphthalic acid (the elemental analysis was cor rect for this structure). Elemental analyses of all new phthalic anhydrides (and 3"iso- propylphthalic acid) have been obtained and are recorded in Table XII. Table XII Phthalic Analysis Theoretical Anhydride % Q ^ H % C ^ H 3-n"Propyl” 69,31 5.20 69 .if 9 5.30 3-Isopropyl-(acid) 63.62 ^.77 63*if5 5 .8I 3-Methyl-6-ethyl- 69,69 5.^3 69.15 5.30 3-Kethyl-6-n-propyl- 70.39 5.91 70.57 5-93 Tetramethyl- 70.62 6.07 70.57 5.93 Melting points of all phthalic anhydrides (and 3-isopropyl- phthalic acid) prepared are given in Table XIII. Neutral equivalents of these compounds were determined as a measure of purity of product -131- in a number of experiments. In addition, neutral equivalents were determined on recrystallized samples to be used for melting point and elemental analysis purposes (values were always within 1-3^ of the calculated ones)# It is not deemed necessary to record those values in this experimental section. Table XIII ______Melting Points of Phthalic Anhydrides______Phthalic Melting Points Melting Point 3-Methyl- 117-118" 112.5-lll;“(ref.l:l) 3-Sthyl- 100-1j01.5" 101-102" (ref .1:2) 3-n-Propyl- 57-5 8 " 3-Jsopropyl-(acid) 137-138" 3, 6-Dimethyl- Ib2-lli2.3" 11:2-11:3“ (ref .37) 3“Methyl-6-ethyl- 111-3.11.^" 3-Methyl-6-n-propyl- Ij9 . 5 - 5 0 . 5 ‘ Tetramethyl- 261:.2-267" ^ Anhydrides were recrystallized from a warm benzene-petroleum ether (60-fl0 ®) mixture (]/l volume ratio), the acid from hot water. Exceptions: 3-n-propyl- and 3-methyl-6-n-propyl-com- pounds from ether. ~ Quantities of Anhydrides Prepared ^ 3-Metl^lphthalic anhydride 2112 g. 3-Ethylphthalic anhydride U92 g. 3-n-Propylphthalic anhydride U99 g, 3-Isopropylphthalic acid 28 g. 3,6-DlJmethylphthalic anhydride 2736 g. ^ Combined material from aromatizatitm experiments (purified) and additional preparations (crude). 41 M.S. Nerwaan, J. Am. Chem. Soc., 6 ^, 1537 (19UD. 42 E.D. Parker, J. Am. Chem. Soc., J2, 21^1 (1950). -132- Quantities of Anhydrides Prepared^ (Continued) 3“Hfithyl-6-ethylphthalic anhydride 399 g« 3-Methyl-6-n-propylphthalic anhydride iiO$ g* Tetramethylphthalio anhydride 5 g» Combined material from aromatization experiments (purified) and additional preparations (crude) . Other Aromatisation Experiments «— In addition to the tvo aromati zation procedures ifhich were just described, a large number of others were tried. The experiments which were carried out are recorded in Table VIII* These all proved to be unsuccessful. Since adduct and acid quantities, solvent (if necessary) and conditions have been outiiji«;d carefully in this table, no one example is given. However, a general description of the workup procedure is given; After the solvent, adduct and catalyst were mixed, the mixture was treated in one of three ways |l) allowed to stand at room temper ature, 2 ) heated on a steam bath, 3 ) heated at refluj^ } the solution then was poured into cold water and the solid (or oil) obtained was isolated and treated with a dilute sodium hydroxide solution. The solution then was decolorized (if possible), acidified, and filtered. The filtrate was extracted with hot benzene, as was the residue, and th e combined solutions were cooled at ^-10* for 2h hrs. in order to effect the precipitation of any substituted phthalic anhydride (or acid) which might be present. Reduction of 3-Alkyl- and 3~Methy1-6-alkyIphthallc Anhydrides ^ Hydrocarbons Preparation of Phthalate Esters» — 1. Methyl esters; Dimethyl 3,6-dimethylphthalate Dimethyl 3-methyl-6-ethyIphthalate Dime thyl 3-niQthyl-6-n-propylphth alate These three compounds were prepared according to the procedure used by Geissman and Morris*® to obtain dimethyl naphthalate. This method proved to be extremely valuable because diethyl 3“fflethyl-6-ethyl- phthalate and diethyl 3“methyl-6-n-propylphthalate could not be pre pared (possibly because of steric hindranooL A preparation of dimethyl 3,6-diaethylphthalate is described below; A solution of 53 g« (0*3 mole) of 3,6-dimethylphthalic anhydride in 335 ml* of a solution of 112 g* (2 moles) of potassium hydroxide in 1 1. of methanol was placed in a 3-1* round-bottomed, three-necked flask equipped with a motor-driven Hershberg-type stirrer and two addition funnels* Into one of the funnels was introduced the re mainder of the potassium hydroxide solution and into the other 18? ml* (2 moles) of dimethyl sulfate* The two reagents were run in simultaneously at a rate of 1;:1 with stirring and cold-water cooling* After the reagents had been added, the potassium methoxysulfate was collected on a Buchner funnel and washed with 200 ml* of methanol* The combined filtrate and wash liquid was concentrated in vacuo to one-third volume and poured into water. The milky suspension was 43 T.A. Geissman and L* Morris, J. Am* Chem. Soc., 66, 718 (19i;ii). -133- -13L - slialven v.dth 2 200-ml. portion? of ether, the aqueous layer v;as ois- cai’ded, and the ether solution was dried over anhydrous sodium sul fate. The solvent was removed and the oroduct was distilled from, a Claisen flask under reduced pressure to give [j6,2 g. (0.21 mole, 70^) of diiiiethyl 3,6-dimethylphthalate, b.p, Hi2-Ui3“/p mm., m.p. 77-78% Quantities and Yields of Esters Prepared Dimethyl 3, 6-dime thylphth alate i^o.2 g® (70% yield) Dijnethyl 3-methyl-o- e thylphth alate 302*0 g. (60, 62^ yields) Dimethyl 3“methyl-6-n-propylphthalate 300*0 g* (58, 60^ yields) The physical properties and elemental analyses of these com pounds are recorded in Tables XIV and XV. 2* Ethyl esters: Diethyl 3-methylphthalate Diethyl 3-ethylphthalate Diethyl 3-n-propylphthalate Diethyl 3,6-dimethylphthalate These esters were prepared according to a general procedure, i-diich is exemplified below: Two hundred and sixty-four grams (l.ii moles) of 3-n-propylphthalic anhydride, 2 1 * of absolute ethanol, 200 ml* of dry benzene aid 20 ml, of concentrated sulfuric acid were mixed in a 3 -1 . ^ngle-necked flask equipped with an electrically heated mantle and a 1,6 x 100 cm. packed column (fitted with an azeotropic head)* The mixture was heated at reflux until water no longer passed into the receiver an -135” its benzene azeotrope. The mixture then was cooled and poured into water. The benzene layer was separated, the aqueous layer was ex tracted with 2 200-ml. portions of benzene, and the combined benzene material (layer and extracts) was washed with water and dried by per colation through a column of anhydrous sodium sulfate. The solvent was removed and the product was distilled under reduced pressure through column B to give 29h g. (1.1 moles, 80$) of diethyl 3-n-pro- pylphthalate, b.p. 123-125"/0.7 mm., Dg° 1.5003). Quantities and Yields of Ester Prepared Diethyl 3-methylphthalate 1336 g. (82-85$ yields) Diethyl 3-ethylphthalate U80 g. (75“78$ yields) Dietl^l 3-n-propylphthalate g. (75$, 8o$ yields) Diethyl 3,6-dimethylphthalate 2026 g. (72-76$ yields) The physical properties and elemental analyses of these compounds are recorded in Tables XIV and XV. 3. Butyl esters: Dibutyl 3-methylphthalate Dibutyl 3, 6-dime thylphth alate These compounds were prepared in essentially the same manner as were the ethyl esters. However, benzene was not required as the n- butyl alcohol served as the azeotroping agent. No description of the procedure is given. Five hundred and ten grams of dibutyl 3-methylphth alate and 6Ü0 g. of dibutyl 3,6— dimethylphthalate were prepared in yields of -136- 085É and 82%, respectively, 'i*he physical properties and elemental analyses of these esters are recorded in Tables XIV and XV. Table XIV Physical Properties of Phtil alate Esters Boiling Melting Refractive Esters Point Point Index (ngo) Diethyl 3-methylphthalate 155.2V 5 mm. 1.501:3 Dibutyl 3-methylphthalate 186°/L mm. 1.1:956 Diethyl'3-ethylphthalate 11:1°/1.5 mm. 1.5028 Diethyl 3-n-propylphthalate I22-I2 5 V 0.5 mm. 1.5003 Dimethyl 3,6-diinethylphthalate 31:2-11:375 mm. 77-78.6° Diethyl 3,6-dimethylphthalate 15U.5-155% mm. 1.5029 Dibutyl 3,6-dimethylphthalate l65-l66“/0.5 mm. 1.1:95U Dimethyl 3-methyl-6-e thyl- phthalate lliO-3l:l°/2 mm. 1.5lUi Dimethyl 3-methyl-6-n-propyl- phthalate 139-ll:0°/0.9 mm. 1.5002 Table XV Elemental Analyses of Mi th alate Esters Esters Analyses Theoretical % C % H % 0 % H Diethyl 3-methylphthalate 65.31: 7.28 66.06 6.83 Dibutyl 3-methylphthalate 69.36 8.36 69.83 8,27 Diethyl 3-ethylphthalate 67.1:8 7.30 67.16 7.25 Diethyl 3-n-propylphthalate 68.18 7.61: 68.18 7 .6a Dimethyl 3,6-dimethylphthalate 61.62 6.11 61:.85 6,35 Diethyl 3,6-dimethylphthalate 67.05 7.28 67.18 7.25 Dibutyl 3,6-dimethylphthalate 70.80 8.62 70.56 8.55 Dimethyl 3-methyl-6-ethylphthalate 66.30 6.97 66.06 6,83 Dimethyl 3-methyl-6-n-propylphthalate 67.12 7.W: 67.18 7.25 Copper Chromite Method»— The catalytic reduction of diethyl or dibutyl 3j6-dimethylphthalate in the presence of a copper chromite catalyst yielded only very small quantities of prehnitene. Catalyst -137- quantities as large as 30% by freight were used, and reductions were carried out both in the absence of solvent and in dioxane or methy1- cyclohexane. Temperatures were varied from 250“ to 300“ in conjunc tion with pressures of 200 to 300 atmospheres. Stepwise reduction— ester to d i d to hydrocarbon— was tried, but without success. A description of an unsuccessful experiment is jjiven: > One hundred and fifty-five grams (0.6 mole, 150 ml.) of diettyl 3,6-dimethylphthalate (distilled from Raney nickel) was heated at 250“ in a rocking 300-ml. stainless-steel bomb in the presence of 20 g. (12% by weight) of a commercial copper chromite catalyst (Harshaw Nickel Co*^ composition; Cu04t2%, Cr203-li5^j BaO-10%) under a hydro gen pressure of 310-330 atm* (obtained with the aid of a booster pump). No solvent was used. In 2h hrs. the absorption of hydrogen ceased (only one-fourth of the theoretical quantity had been taken up). Heating was discontinued, the bomb was cooled, hydrogen was vented, and the reduction mixture was removed. The catalyst then was filtered and the filtrate was poured into water. The mixture was shaken with 350 ml. of ether, the aqueous layer was discarded, and the ether layer was dried over anhydrous sodium sulfate. Solvent was removed and the product was distilled frcm a Claisen flask under reduced pressure. No prehnitene was obtained. A UO g. residue of white, crystalline material remained. This solid was identified as 3,6-dime thylphth alic -.138“ anhydride: m.p. iLl.S-UjG.S" (recryst.nixed melting point with an authentic sample * The fractions collected durijig the distillation (62.5 g.) were combined and treated as follows; 1) The combined material was dissolved in a solution of 20 g. of potassium hydroxide in 200 ml* cf absolute ethanol and the mixture was refluxed for three hours, then poured Into water. Two hundred milliliters of ether was added, the mixture was shaken in a separatory funnel and the aqueous and organic layers were separated. Acidifica tion of the aqueous layer gave $ g. of white, ciystalline material which melted sharply at 88.8-89.2" after one recrystallization. Its infrared spectrum contained a strong band at 17^2 cm. ^ which was interpreted as the carbonyl stretching band of an^^, ^ -unsaturated lactone. The only logical structure of the above solid was concluded therefore to be 3,6-dimethylphthalide. An elemental analysis of this compound supported this assumption. Elemental Analysis; C, 7li*31j H, 6.27 (calculated for 3,6-dimethyl- phthalide, ; C, 7^*CUj H, 6*22). 2) The organic layer from the above extraction was dried over sodium sulfate, the solvent was removed, and the product was distilled slowly under reduced pressure through a short packed column. Only one compound of the mixture could be separated. Ten grams of a liquid distilled at llL"/20 mm. (ng° 1.533?)» Its infrared spectrum con tained a very strong band at 1080 cm. ^ which was interpreted as the -139- C-O-C stretching band of a cyclic ether. The compound was assumed, therefore, to be 3,ü-benzo-?,<-dlhydrofuran. Elemental analysis of the structure supported this assumption. Elemental Analysis; C, 80.90%; H, 8,U9%(calculated for 3,L-benzo- 2,5-dihydrofuran, CioH^gO; C, 8l.cU%; H, 8.11^. The catalytic reduction of diethyl or dibutyl 3-methylphthalate took place to give hemimellitene only when they were heated for long periods under high hydrogen pressures and at high temperatures. A brief outline of a successful preparation is given: One-hundred and forty-six grams (0.5 moles) of dibutyl 3-methyl phthalate was hydrogenated at 2?5* and 300 atm. in the presence of 62 g. (30% by weight) of copper chromite (composition: CuO-39%, CrgOg-liT^, BaO-11%). Twenty-four hours were required. Distillation of the product through column B gave ii6.8 g. (0.39 mole, 78%) of hemimellitene, b.p. 176-176.8“/7U7 mm., n*® 1.5liiO (reported^: b.p, 177.07V760 mm., ngo 1.5139). Reduction of Phthalic Anhydrides to Hydrocarbons .— 3.6-Pimethvl- phthalj.c anhydride was reduced to prehnitene by the following steps: 1. A 2-1. three-necked, round-bottomed flask was equipped with a motor-driven Hershberg-type stirrer, a separatory funnel, an elec trically heated mantle, and a Soxhlet extractor. The Soxhlet thimble was packed with Ui g. (0.25 moles) of powdered 3,6-dimethylphthalic anhydride, the apparatus was flushed thoroughly with nitrogen, wnd a tube was led from the Soxhlet outlet to a nitrogen-filled reservoir -lilO- (1 gal. bottle). One liter of dry ether was placed in the flask and 10.5 g. (0.27 mole) of lithiuE aluminum hydride were introduced through a powder funnel. Stirring was commenced, and the mixture was refluxed until the contents of the thimble were exhausted (6 hours). Tlie flask then was cooled with Dry Ice and a 20$ hydrochloric acid solution was introduced through the funnel until the lithium and aluminum salts settled in the form of sma]-l pellets. The ether solu tion was decanted, the pellets were washed with 2 lOO-ml. portions of ether, and the combined material (solution and washings) was dried over anhydrous sodium sulfate. The solvent was removed and the prod uct was recrystallized from hot petroleum ether (60-ÜO" ) to give 36 g. of a white crystalline solid, m.p. 68-89®. The infrared spectrum of this compound Indicated that it was not a diol. A comparison of the spectrum with that obtained earlier for 3,6-dimethylphthalide proved that the substances were identical. A mixed melting point of the ccjqxjunds showed no depression. A yield of 90$ can therefore be calculated for the reduction step, 2. When the above reduction procedure was carried out on 3,6-dl- methylphthalide (36 g* 0.22 mole), with U.6 g. (0.12 mole) of lithium aluminum hydride, 32 g. (0.2 mole, 90$) of 3,6-dimethyl-o-xylylenol (m.p. 70-70 .5 * ) were obtained. Elemental Analysis: C, 72.5h$; H, 6.57$ (calculated for 3,6-dimethyl- o-xylylenol CigHi^Og: C, 72.26$; H, 8.U9$). -iLl- The above two steps were repeated to give an additional azoimt of diol (2d g«; step 1, 0?^ yieldj step 2, 90% yield). 3. Sixty grams (0.38 mole) of 3,6-dlmethyl-o-xylylenol,6d g. (0.214 mole) of purified phosphorus tribronri.de, and 200 ml. of dry benzene were refluxed for three hours. The benzene solution was decanted frcm the phosphoric acid, washed with cold water, and dried over anhydrous sodium sulfate. The solvent was removed and the prod uct was recryatallized frcm hot petroleum ether (60-80®) to give 105 g. (0.36 mole, 95%) of 3 ,6-dimethyl-o-xylylene bromide, m.p. 9?.5~1 00 .5 "« No elemental analysis of this compound was obtained, but Buchta and Loew*^ later reported a melting point of 100“ • I4. The final reduction step was carried out in the same manner as the first and second. Fifty-two grams (0.18 mole) of 3,6-dimethyl- o-xylylene bromide was reduced by a mixture of 2.2 g. (0*27 mole) of lithium hydride and 3 *U g» (0.09 g. ) of lithium aluminum hydride. Bae products of two runs of equal size were combined and dis tilled through column A to give 36 g. (0.27 mole, l5% ) of prehnitene, b.p. 209“/7U1 mm., 1.5202 (reported'*: b.p. 20U.63“/760 mm., ng° 1.5202). 3-Methylphthalic anhydride was reduced directly to the diol by lithium aluminum hydride in a single small run. The preparation Is outlined here. Sixteen grams (O.l nK>lo) of the anhydride was reduced by h g. (0.11 mole) of lithium alum1 num hydride in a 1-1 . flask in the same 44 E. Buchta and G. Loew, Ann., 5 9 7 , 123 (1956), -Ib2- maimer as described on p. UiO* The product was recryatallized from hot petroleum ether (35~^45>" ) to give 13 g. (O.O8Ç mole, 6^^) of 3-aethyl-o-xylylenol, m.p. ^0^1*. elemental Analysis; C, 70.95^> H, 8.01% (calculated for methyl-o- xylylenol, CgHigO^: C, 71.02%; H, 7.95%). Reduction of Phthalate Esters to Hydrocarbons.— Reduction of dimethyl and diethyl phthalates to diols by lithium aluminum hydride proved to be an excellent method* Reductions were carried out rapidly and conveniently and on a reasonable scale, in sharp contrast to the slow, limited method described in the last section. A description of a general procedure is given. A S-1. round-bottomed, three-necked flask was equipped with a motor-driven Hershberg-type stirrer, an addition funnel, and a largo Liebig condenser from which was led a tube to a large nitrogen reser voir (1 gal. bottle). Ihe flask and bottle were flushed with nitrogen and the calculated amount of lithium aluminum hydride was placed in the flask (allowing a 10% excess)• One and one-half liters of dry ether per mole of reducing agent then was added. Stirring was begun and the solution of ester was introduced dropwise, with cooling (this was effected best and most safely hy a pan of powdered Dry Ice). When the addition was completed the mixture was stirred for one hour longei) then was hydrolyzed by adding a 20% solution of hydrochloric acid (with thorou^ cooling) until the lithium and aluminum salts settled in the form of small pellets. The ether solution was decanted, the pellets were washed with ether and the combined material (solution and washings ) was dried over sodium sulfate. Solvent was removed and the product was either distilled or recrystallized from petroleum ether (6O-8O®) (low melting diols were best purified by distillation). Quantities arid Yields of Diols Prepared 3-Methyl-o-xylylenol 886 g. (88-915É yields) 3-Sthyl-o-xylylenol 282 g. (86$, 88$ yields) 3-n-Propyl-o-xylylenol 306 g. (89$, 90$ yields) 3,6-Dimethyl-p-xylylenol 1030 g. (87-90$ yields) 3-Methyl-6-ethyl-o-zylylenol 211 g. (90$ yield) 3-Methyl-6-n-propyl-o-xylylenol 1^9 g. (05$ yield) The elemental analyses and physical properties of the diols are recorded in Tables XVI and XVII# Table XVI ______Elemental Analyses of_Diols______Analyses Theoretical ______Diol______$ C $ K ______$ C $ H 3-Methyl-o-aylylenol 70.95 8.01 71.02 7.95 3-Ethyl-o-xylylenol 72.23 B.Uî 72.26 8.U9 3-n-Propyl-o-xyly^lenol 73*12 8.38 73.33 8.95 3,?-Dime thyl-o-xylylenol 72.51» 8.57 72.26 8.i»9 3-Methyl-6-et%l-£-3cylylenol 73.35 8.95 73.33 8.95 3-Methyl-6-n-propyl-o-xylylenol 7U«U9 9*35 ?U.l6 9.3U Table XVII Melting Boiling Refractive Diol Point Point Index (nj>=) 3-Me thyl-o-xylylenol 50-51* 160-l6l'/2*5mm* 3-Ethyl-o-xy ly leno 1 32-33* 151 .8-152.5 "/ approx* 0 .7mm. 3-n-Propyl-o-xylylenol 152-153 */ 1.51<38 approx* 1mm. 3,6-Dime thyl-o-xylylenol 70-70.5*^ 3-Me thy 1-6- ethy 1-o-xyly len ol 78-7 8 .5 * 3-Methyl-6-n-propyl-o-xylylenol ^ Reported m.p. 70* (ref. Mi ). Reduction of the diols to the hydrocarbons was effected con veniently by an extension^® of the Birch'*® procedure* A general procedure is described for a one-mole run* A round-bottomed, three-necked flask was equipped and in sulated as described on p* (reduction of 2-vinylfuran)* Two liters of liquid annnonla was fed in from a commercial tank and p6*6 g. (1^.2 moles) of freshly cut sodium was introduced rapidly in it-5 g* pieces. Stirring was commenced and when the sodium had dissolved (approx. one-half hour), a solution of one mole of diol and 6U g* (2 moles) of methanol in 1 liter of dry ether was added dropwise at a rate sufficient to cause a steady reflux of ammonia* When the addition was completed (usually within one hour), 227 g* (i|*2 moles) of powdered ammonium chloride was introduced immediately from an 46 D*C. Rowlands, Doctoral Dissertation, The Ohio State University, 1952. 46 A.J* Birch, J* Chem* Soc*, 19kS* 809* ”lü5“ Erlenmeyer flask (90® bent neck) which was attached tn the flask by a length of Gooch tubing. Water then was added until ammonia no longer refluxed; the aqueous layer was siphoned and discarded and the ether solution was washed with water and dried over anhydrous sodium sulfate. Solvent was removed and the pixiduct was distilled under reduced pressure Inception: 1,2,3-trimethylbenzene (hemimellitene) was distilled under ordinary p r e s s u r e ] • Quantities and Yields of Hydrocarbons 1.2.3-Trimethylbenzene 62k g* (89-92% yields) 1.2-Dlmethyl-3-ethylbenzene 20$ g. (89%, 90% yields) 1.2-Dimethyl-3-u-propylbenzene 226 g. (90% yields) 1,2,3,U-Tetramothylbenzene 750 g. (88-90% yields) 1.2.3-Tri«ethyl-3-ethylben2ene i9l g. (85% yield) 1.2.3-Trimethyl-3-n-propylbenzene 130 g. (80% yield) Elemental analyses of three new hydrocarbons are given in Table XVIII• Table XVIII ______Elemental Analyses of New Hydrocarbons______Analyses Theoretical ______Hydrocarbon______%C %H %C %H 1.2-Dime thyl-3-n-propylbenzene 89.19 10.97 89.lU 10^38 1.2.3-Trimethyl-3-athylbenzene 89.12 10.98 8 9 .Ih 10.68 1.2.3-Trime thyl-3.g-propylbenzene 88.71$ 11.36 88,78 11.18 Boiling points, refractive indices, densities, freezing points, Tfo and cryoscopic constant A values, and purities of w11 hydrocarbons prepared are recorded in Table XIX. Table XIX Hydrocarbon b.p. f.p.(m.p. ) d20 Tfo A Purity 4 (mole %) 1,2,3-Trimethylbenzene 176.02“ -25 .62“ 0 .69% 1.5139 99.61 (Literature^ ) 176.0%“ -25.375“ 0.691:38 1.51393 -25.375“ 0.0161: 1,2-Dimethyl-3-ethyl- 193.89“ -1:9.686“ 0.3922 1.5117 0.029 99.11 benzene (129“/I2$mm.) (Literature* ) 193.91“ -1:9.5“ 0.3921 1.5117 -1:9.38“ 1,2, 3, li-Tetrame thyl- 20L.S3“ - 6 .26“ 0.901:7 1.5202 99.9 I benzene (90“/l^mm. ) -O.I (Literature * ) 205.oU“ - 6.25“ 0.9052 1.5203 - 6.25“ 0.01901 T 1, 2-Dimethyl-3"S," 211.76“ -26.90 “ 0.3828 1.5068 -26.521:“ 0.026 99. So propylbenzene (1L1.^“/ 1 1 W . ) - 0 .06“ 1,2, 3-Trimetliyl- 220.95“ -51.01“ 0 . 9 0 1 5 1.5176 - 5 0 . % “ 0.031: 9 8 . 0 3 ii-athylbenzene ( % 2 “/83mm. ) 1 0 .08“ 1,2,3f-Trimethyl- 237.0 -18.22“ 0 . 3 9 3 3 1.5130 -18.205 0.035 99.93 U-n-propylbenz ene - 0.1“ - 0 . 0 0 1 “ - 0 . 0 3 (0l;9“/7Omm. ) * F. D. Rossini, ^ cO.., Selected Values of Properties of Hydrocarbons and Related Compounds. American Institute Research Project Ijli, The Carnegie Press, Carnegie Institute of Technology, Pittsburgh, Pa., 19^2, Tables 5z, 2liz, ^a, ILa. SUMÎ.ÎARY 1. A thorough investigation has been carried out in order to determine the scope of a method of preparing 2-dimethyi-3-alkyl- benzenes and 1,2,3“trimethyl-ii-alkylbenzenes from 2-alkylfurans and 2-methyl-5-alkylfurans. The research program is summarized below; a. Four 2-alkylfurans and two 2-methyl-5-alkylfurans ware pre pared on a intermediate scale. Modifications of known procedures were made when necessary and a new method of reducing a furylolefin to an alkylfuran was applied in two cases. In addition, two other furans were obtained from commercial sources. b. The Diels-Alder adducts of the substituted furans and maleic anhydride were prepared (exception; a 2-tert-alkylfuran would not condense with the dienophile). In order to expedite the formation of the adducts, and to insure the formation of finely divided material (for aromatization purposes) in certain cases, several modifications of the usual Diels-Alder condensation procedure were made. c. Numerous experiments were carried out in order to determine the best method of converting the adducts to the 3-alkylphthalic anhydrides and 3-methy1-6-alkylphthalic anhydrides. A satisfactory method of aromatization was developed from one of several procedures reported in the literature. Seven anhydrides were prepared in this m anner. Only the adduct of a sec-alkylfuran-maleic anhydride could not be aromatized in an acceptable yield by this procedure. d. An investigation of various methods of carrying out the reductions of the anhydrides to the hydrocarbons was made. The best -lli7” -118- proceduro was found to t>3 one v,tiich entailed tho preparation of the esters, rociuction of the esters to diols and finally, reduction of the diols to the hydrocarbons. In the last step a now extension of the Eirch reduction was applied. Six hydrocarbons vfcre prepared from the anhydrides by this three-step procedure. e. Physical properties (densities, boiling points, freezing points and refractive indices ) and purities of all hydrocarbons ’'/ere determined. 2. In addition to the furans required for tiie preparations of 1,2-dimethyl-3-e.lkylbenzenes and 1,2,3"trimethyl-i}-alkylbenzenes, three 2-furylcycloalkanes (two of them new) and four miscellaneous 2-substituted furans were synthesized and their maleic anhydride adducts were prepared. However, with a single exception, the addends could not be aromatized by the procedure 'vdiich had given satisfactory results with the adducts of the 2-alkyl- and 2-methyl-^-alkylfurans. The two new 2-furylcycloalksnes were prepared according to a pro cedure which incorporated two new reactions of fur an compounds. In order to prepare another 2-substituted fur an, a modification of a known procedure had to be made. In the course of this research (outlined here as 1. and 2.) l7 new compounds were prepared and characterized by physical properties and elenental analyses (carbon and hydrogen). AUTOBIOGMPW I, Earl Phillip f.foore, Jr., was born in Detroit, l'ichigan, Jarrary, 16, 1927* î'y secondary school education was obtained ab Mackenzie High School in Detroit, ?,{ichigan. In 19li5 I entered active duty with the United States Navy. Upon release from active duty with the Navy in 19ii6, I entered the University of Miami, Florida, where I completed the requirements for the degree of Bachelor of Science in 1950 and the requirements for the degree of Master of Science in 1952. I matriculated at the Ohio State University in September, 1952. I held a Research Assistantship with the American Petroleum Institute Research Project h5 from 1 952 to 1955» I vras appointed a General Motors Fellow in Chemistry from January, 1959, to January, 1 956. I was appointed a du Pont Fellow in Chemistry from January, 1956, to July, 1956. I held a Research Assistantship with the American Petroleum Institute Research Project from July, 1956, until completion of the re quirements for the degree of Doctor of Philosophy. -U;9-