w

H
l w

W

i

H
M

l

l

1'

w
n‘

I”
it)

I

 

THE ATTEMPTED FORMATlON
OF 8-ALKYL CAFFEINES
BY THE PYROLYSIS OF CAFFEINE
ETHERS \X‘ITH ALKYL ETHERS
1%. THE PREPARATION AND

MOLECULAR REARRANCEMENT
OF TWO ETHERS OF CAFFEINE

Thesis fer the Degree of M. S.
7/ ‘- x n \
fins-mum}. D. Bacon

1936

I THE ATTEEIPTED Fommmmz 0F 8-11LKYL Cpmzmms
BY T1113 PYROLYSIS 01? 01113112131112: TTZ'JT’S
1111111 ALKYL 13111112113

II TIE?! PREPARATIOET A231) MOLECULAR RBMTEZAUGBEEH

0F T‘C-EO ETEBERS 013‘ CAFFZ'I INE

Thesis
Submitted as Part of the Requirements

For the Master of Science Degree

Kenneth D. Bacon

1956

ACKN OWLEDGIJEH T

The author wishes to czpress his sincere
appreciation to Dr. R. C. Huston, whose helpful
suggestions and sound advice made possible the

completion of this work.

331602

CON TIBETS

Page

HISTQRICAL wflun-fl-uuq—naQ—g 3

Emmnmmgz. .. .. .. .. .. .. — - - - ~ .. - - - 13

I
I
t
I
i
I
!
1
t
I
I

SUMILARY MED DISCUSSION .28

BIBLIOGRsPIFI -------'—--- ~------ :51

INTRODUCTION

It was Emil Fixner and.his associates that carried out
the first extensive work on compounds containing the purine
nucleus. His investigations led to the final decision
regarding the structure of uric acid, both.by its synthesis
and by its reactions. He synthesized all of the mono, di,
and trimethyl derivatives of uric acid that are possible by
the theory of structure expressed by the formula suggested
by Medians in 1875, and proven by Fixher himself later.
Finally he converted.uric acid into purine, which.contains the

nucleus common to caffeine and related compounds.

11- «inc ”-8.3
7
O“ .N-H H- h “H
E I :30 is 5 Kean-H
Hz. a uN-IH 5'- -9
4
Uric Acid Purine

1 , 1
‘ 1‘ . C h H
CH3 N ~ 0 - N
Caffeine
Caffeine is of interest to study, because of its occur-

once in the two most common beverages, tea and coffee, and
its physiological effect when 'taken internally. It has a
pronounced effect on the nervous system, it stimulates the
heart action and it is a strong diuretic. It is possible, by
comparing the physiological action of theobrcmine, thesphyl~
line and caffeine, to show that their different effects on

2.

the system.are due to introduction of alkyl groups, in this
case methyl, into the purine nucleus.

It is not caffeine itself that interests us in this work,
but rather certain derivatives of caffeine, particularly the
8-cthers of caffeine. The purpose of this thesis will be to
accomplish the following aims:

1. The attempted formation of 8-alkyl caffeines by .
process reported by Winston F. Allen from.this laboratory.

2. The preparation and preperties of certain.new
caffeine others. ‘

3. The attempted molecular rearrangenent of those pron

pared others.

3.

HIST RICAL

A. Early work on caffeine-others.

In the course of Fischer's extensive, experimental work
on compounds of the purine typo, he prepared only two of the
8—ethers of caffeine, Bemethoxy, and B—othoxy-caffeine.(1)
Fischer prepared these others by heating 8~chloro, or 8—brcmo-
caffeine in a methyl or ethyl alcohol solution containing an
excess of sodium or potassium alcoholate. The alcoholate was
made by adding either the metal or the hydroxide to the also-
hol. Later W. Wislaconuscz), and H. Biltz(5) repeated this
work. ' d

(4.)

A. Baumann wont farther than this in preparing fif~
teen 8-substituted~phonyl others of caffeine. His method
consisted in.heating 8-chlorocaffeine with.a substituted

phenol in, the presence of an equivalent amount of potassium.
hydroxide. He was able to carry out the reaction in aqueous
solution with refluxing and under pressure at high temperatures.

In some cases he found that.xylem3sas more satisfactory as a

solvent, in place of the water.

B. The work of Huston and Allen‘s).
1. General methods of preparation of the B—others of
caffeine.
In 1932 the work on the preparation of sixteen
8-othors of caffeine was completed in this laboratory. Many

of those others were new compounds. The use of both B—chloro

4.

and 8~bromo-caffeine was employed, with B—chloro-caffeins
being preferred in most cases due to its ease and economy of
preparation.

Clean metallic sodium was cut into small pieces
and added to enough absolute alcohol to permit the resulting
alcoholate to readily dissolve. The pr0portion was generally
about 1 gram of sodium to G mls. of the alcohol. When the
sodium had entirely disappeared an equivalent amount of the
B-ohloro-caffeine was added and shaken well. This mixture
was refluxed on a steam or oil bath for one-half to five
hours, depending upon the speed of the reaction. The sodium
chloride was removed from the hot solution by immediate
filtration. Some of the lower alkyl caffeine others crystal-
lized out of the alcohol solution very satisfactorily when
cooled. However, some of the higher others would not give
good crystallization unless the solution was concentrated to
a smaller volume. This was done by vacmlm distillation, in
order to keep the temperature at a minimum. In some cases it
was necessary to add an equal volume of water to the concom-
trated solution to obtain complete crystallization. Recrys-
tallization was best carried out from 40 to 50 per cent ethyl
alcohol. Each other seemed to offer its own particular
problem where the higher alkyl groups were concerned. Those
others which Allen prepared will be taken up in detail later.

2. Reactions of the Brothers of caffeine.
(a) Formation of hydroxy caffeine or 1-3-7 trimethyl‘
uric acid(5 1

5.

It was found that the 8—ethers of caffeine would
form hydroxyc-caffeine by two methods:

(1) By heating the other in a closed tube at
high temperatures. Using this method it was possible to
obtain hydroxy-caffeine from all of the others to some extent
except 8-phenyloxy-caffeine and 8~para-hydroxy-phenyloxy-
caffeine. 8+.Iethoxy, 8~benzyloxy, and 8-phenyl-éethoxy-eaffeine
gave very small amounts of hydroxy-wcaffeine.

(2). By heating the other with dilute hydroe
chlcric acidw). Hydroxy-oaffeine was formed by the treatment
of the caffeine other with five parts by weight of 10 per cent
hydrochloric acid in a flask, heating the mixture on a steam
bath for twenty to thirty minutes. The other gradually went
into solution, and when the reaction had gone well toward
completion the hydroxy-caffeine separated out while the
solution was still hot. Upon cooling long slender needles in
clusters formed. These crystals were filtered off, the fil-
trate made almcst neutral, and concentrated down to nearly
one-fifth of the original volume, thus a second crap of
crystals were obtained. The entire product of crystals was
recrystallized again from hot HOE or alcohol.

This reaction seemed to be typical of tee
B-alkyl others of caffeine. The basis for this statement lies

in the fact that hydroxy-caffeine was formed from the alkyl

others up to and including n. heptyloxy-caffeine, in some

cases readily. It was further found that the two phenyl alkyl

6.

others of caffeine prepcred by Huston and Allen would form
hydroxy-caffeine in the same manner. _. Phenyloxy and p.
hydroxy~phenyloxy-caffeine would not be converted into hydroxy-
caffeine by any method however.

(b) Molecular rearrangement of the B—ethers of

caffsinehtetra-alkyla-uric-acids(5) .

The second reaction that Huston and Allen found
in the caffeine others was a less general reaction. It con-
sisted of the molecular rearrangement of some of tho B-alkyl
caffeine others to the corresponding alkyl substituted uric
acid. It was reported, however, that onlyfive of the pre~
pared others would give this rearrangement.

In repeating the work of Biltz and his co-work-
are”) on the molecular rearrangement of 8~methoxy and 8-
ethoxy caffeine to tetra methyl, and trinethyl 9-ethyl~uric
acids, it was found that the rearranges-sent would take place
as reported in a closed tube at temperatures up to‘250°c.
Huston and Allen went flirther, however, and found that this
rearrangement would take place better in an Open tube heated
in a paraffin bath.

The general procedure followed was as follows:
From two to five grams of the caffeine other was heated in
either a closed or an Open tube at temperatures from 240° to
305°C. for anywhere from 4 to 18 or more hours. In the course
of the reaction it was found that the tube contents often
formed a solid mass and a readily vaporized liquid after the

heating had been continued for two to three hours. The solid

7.

substance was thought to be hydroxywcaffeine. Varying
amounts of hydrexy-caffeine were formed.with the rearranged
product.

The trimethyl 9—alkyl-uric acids were separated
from hydroxy-caffeine as well as any unchanged caffeine other
by boiling the product with 10 per cent HCl,and then after
evaporating to dryness the residue was dissolved in a small
quantity of hot water. The solution was carefully neutralized
with.barium.hydroxide solution, forming the insoluble barium
salt of the hydroxy-caffeine. Once more the solution.was
evaporated to dryness. The residue was extracted with
several small portions of hot chloroform. 'The soluble tri-
methyl 9-alkyl-uric acid was thus taken.off, leaving the
barium salt of hydroxy-caffeine. After distilling off the
chloroform, the tetra methyl uric acid remained in a very
pure state.

These others which would rearrange follow:

1 - 8—methoxy-caffeine. Up to 95 per cent rearrangement
either in the presence of a small quantity of absolute methyl
alcohol, or without it, in a closed tube, or better in an
Open tube, at temperatures up to 560°. The best conditions
for rearrangement were in an epen tube at 240-2500 for a
period of four hours.

2 ~ B-ethoxy-caffeine. Equal amounts of the dry ethoxyn
caffeine and absolute ethyl alcdhol were heated in an Open
tube at 245-255° for eight hours. These were the conditions

for cptimum.yield of the rearranged product, which amounted

 

8.

to 51.2 per cent of the theoretical;

3 b B—ne prOpyloxyucaffeine. To obtain Optimum.rearrangenent,
this other had to be heated for ten hours at zoo-270° in an
Open tube. At the end of this period the oily liquid residue
was removed with.hot water, clarified with animal charcoal,
and then any hydroxy-caffeine or unchanged ether was removed
by the barium hydroxide method. A 35.5 per cent yield was
obtained.

4 . B-allyloxy~caffeine. Due to the fact that this was an
unsaturated ether, it rearranged readily. By heating in an
cpen tube at 170~185° for four hours, a maximum yield of 53.8
per cent trimothyl 9-allyl-uric acid was obtained. Higher
temperatures or longer heatinr lowered the yield.

5.. Benzyloxyncaffeine. The other was heated in a closed tube
at cod-205° for ten hours for a maximmn rearrangement of 50.0
per cent us the trimethyl 9—benzyl-urio acid.

These five caffeine ethers were the only ones
that Euston and Allen could.cause to rearrange. By using
more drastic heating, over longer periods of time, it was
found that the result was either the formation of more hydroxy-
caffeine, or else decomposition of the ether into unrecognis~
able substances. Thus there seemed to be a temperature at
which there was a maximum of the rearranged product formed,
and at vagich there was a minimum of decomposition products
and hydroxyucaffeine formed.

The five rearranged.products were more soluble

9.

in alcohol and water than were the original other compounds.
All of them;melted at temperatures higher than did the
corresponding others. All except trimethyl 9 benzyl-urio
acid sublimed uhen.heated above their melting points. They
possessed a more bitter taste than the others. In contrast
to the original other compounds, they were very stable toward
hot dilute 301, but very easily decomposed by dilute alkali.
All gave a pronounced murexide reaction.

They report it impossible to bring about a re~
arrangement in the ethers having a phenyl group attached
directly to the oxygen on the eifihth caffeine position, or
having two carbons between the benzene nucleus and that ox'éen.
(c) Formation of the 8-slkyl-caffeines‘15).

In this laboratory.Allen reported on the prepar-
ation of certain of the B-alkylccaffeines by two methods.
Each.procedure consisted in the pyrolysnrof caffeine com»
pounds. They are as follows: i

(l) The conversion of the B-alkyl others of
caffeine to 8—methyl and B—ethyl caffeine by'heating in the
presence of alkyl acid anhydrides. @he probable equation for
the reaction is

033 T a a 0
O a ? a . “\f‘gfl83 + O B (COCH3)2 260 u 270°

OHS-H ~ . a 3-5 Hours 3

CH3 s - - 033

l

o sf E- N\- cn5 + R ~ coocs3 + 002

”.6133

0331a.» ~N

The Gomothyl—caffeine was prepared by seating 2 gr.
anhydrous alkylncaffoinc other with an excess of acetic
anhydride. This was done in a scaled tube at 260 to 270° for
from.4 to 5 hours. After cooling, the dissolved contents of
the tube were washed out with.hot alcohol and evaporated to
dryness on a steam bath. The residue was dissolved in 100
mls. of hot water and then clarified by boiling with animal
charcoal. The filtrate was concentrated to 10-15 mls. and on
cooling, the 8-methyl—caffeine would separate out. The pro-
duct was recrystallized out of 95 per cent alcohol. The
resulting white, thin platetets were then.dricd at room
temperature and finally at 100° for five hours.

8-Ethylucaffeine was prepared using the same pro-
cedure as for Benethy1~caffeine. However, it was necessary
to heat the reaction mixture at a considerably higher temper-
ature to obtain the desired product. As a result of this
high temperature considerable amounts of hydroxyucaffcine were
found. This was separated fron.8-ethyl caffeine by the barium
hydroxide method, precipitating the hydroxyncaffeine as the
barium.salt, and extracting the residue with.hot chlorofrom.
This solvent was removed and finally the product was re-
crystallized from alcohol.

(2) The formation of 8-ethyl-caffeine from hydroxy—

caffeine.

11.

CH5 - a - c = o
o a 0» fi ~ a); on H + o (oc02H5)g a
CH3 - N -»c - N/'
co2 + ca3 - a - c a o
o - é o - N\: CH3 + czas-coon

CH3 - t - g o N"C-CZH5

Anhydrous hydrexy-caffeine was heated with an
excess or prepionic anhydride. The temperature was raised to
560°, at which.time the furnace was shut off. The blackened
contents of the tube were washed out with hot alcohol and
evaporated to dryness. *Thc dissolved product was clarified
by'boiling with animal charcoal in water, from which.after
filtration the B-cthyl-caffeine was crystallized. Allen
reported that a temperature lower than 360° was not sufficient
to convert the hydroxyecaffeine to ethyl-caffeine. Further,
he found.that allowing the reaction to take place in the
inert gas, nitrogen,did not result in an increased yield. In
fact the yield.was decreased.

He concluded then that a temperature of 3500 main»
tained for thirty minutes, or 355° for twenty minutes resulted
in a maximum yield of B-ethyl-caffeino.

(5) Formation of 8-dlkyl-caffeines by pyrolysis of
8-pr0pyloxyucatfeine with alkyl others.

Allen.reported.that by heating n. propyloxywcaffeine
with an excess of n. prepyl other in a closed tube at 500-3050

for eight hours, he obtained Bnethylucaffeine with a 37.5 per

 

.9'5.‘ wilhflerwiuihEr! r.-.~:Iu.n.h‘ 1.W a «N. . . an? (.I‘ ‘

12.

cent yield. Here again he followed the procedure of clarify~
ing the water solution of the dissolved product with animal
charcoal, and then crystallizing fromzwater.

Secondary prepyloxywcaffeine was heated with an
excess n. prepyl other at 295~500° for eight hours. The
contents of the tube were treated as before. After re-
crystallizing out of hot alcohol he reports a 52.5 per cent
yield.

Allen, in his report, says that the mechanism of
this reaction can.be explained on the basis of an intermedi-
ate formation of propionaldehyde. The pyrolysis of an
other to an aldehyde is reported by Hurd‘a) and by Hinschel-

wood and Ashey(9)

II
0
caavocssg 500 czasc = o + 0338

-—‘-->
At this high temperature the aldehyde reacts with the caffeine
other to form the 8-alkyl-caéfeine.
Caffeine - O - R + O 3 C - 02H5.__.4>
Caffeine - CQH5 + B-COOH

There is no report on the attempted formation of
8-81kyl-caffeines, using other others than n. prepyl other.
There was an attempt made to prepare 8 n. butyl-caffeine by
heating hydroxy-caffeine with n. bytyric, and n. valeric
anhydride. Heating was done in a closed tube from zoo-558°.

In each.atterpt some of the hydroxy-caffeine was recovered,

but the remainder had decomposed.

I’IrJI 4.3. at trait-.2! in. HR? 5

R? I: re...

. 3. 'HI...

13.

EXPERIMENTAL

A. The attempted formation of Ballkylacaffeines by the

pyrolysis of caffeine others with alkyl others.

1. Preparation of B-chlorOueaffeine.

8-Chloro~caffeine was prepared by the method of

Fischer and Reese‘l). Fifty grams of caffeine, dried at 120°
for at least five hours, was dissolved in four hundred grams
of anhydrous chloroform. This was placed in a flask fitted
with.mercury seal stirrer, reflux condenser with calcium
ahloride tube and a delivery tube to bottom of flask. Chlorine
gas, dried by passing through concentrated sulfuric acid, and
finally phosphorous pentcxide, was bubbled through the mixture
at the boiling point of the solvent, while stirring. The
solution first became very cloudy, but after about thirty
mdnutes or so it cleared up, and at that point the action was
complete. The solvent was distilled off on a water‘baflh,
leaning behind the 8-chlcro-caffeine as a white precipitate.
This was digested with a small quantity of water, and then
allowed to air dry. Finally it was dried at 120° for five
hours. Before using any of the chloroucaffeine it was always

dried at that temperature, to give an anhydrous compound.

A..._‘_4 _. _._ ..._ _._

 

 

Trial 7 *#A_ M__ L #17 -Aji A _5

Material - grs. 50 “V 50 ' 50
Actual Yield - grs. 46 44 51
Theoretical Yield - grs. 58 58 58
Per cent 79.4 76 88
Melting Pt. determdned «186-188 186-187 187

" (Fischer) 188°
Nitrogen a determined 24.27%

calculated 24 .05%

.‘a

m’,~...t...\xp4a\f.-.I¢wu. (Ll-{333. 4 .

ti .v.r.....%fia...

14.

2. The preparation of 8~pr0pyloxyocaffeinm

The espr0pyloxy~caffeine was prepared after the
manner of Huston and Allenw). The prepyl alcohol was dried
with a few pieces of sodium, and then distilled, the fraction
boiling from 97-98° being used. Two gar-ans of freshly cut
sodium were dissolved in one hundred grams of the alcohol.

To this sodium alcoholate solution, twenty grams of anhydrous
Bechloro-eaffeine were added. The mixture was refluxed on an
oil bath, with stirring, for one hour.

While the reacted solution was hot it was filtered to
remove the salt, washing the filter with a small quantity of
absolute alcohol. On cooling the prepyloxy-caffeine crystal-
lised out. This first crcp was filtered and the filtrate
concentrated down to obtain a maximum yield of the other.

The 8 propylcxyocaffeine was recrystallized from 95 per cent

 

 

alcohol.

Trial M l A A 2

F. prcpyl alcohol . grs. 105 100
Sodium .. grs. 2 2
8~Chloro~caffeine 20 20
Yield ~ actual 18.2 17.6
Yield - theoretical 22.1 22.1
Yield - percentage 82.1 80.0
as. - determined lee-130° ‘

M.P. — correct. . 129.5/1so.5°

3. Preparation of alkyl others.
a. Preparation of di-n. prepyl etherul) .
N. prepyl other was prepared by the action of

sulfuric acid on n. prepyl alcohol. Thirty grams of redi—

MIC-ink: 7th.: r-..u.rnm.. Ni‘ ..

15.

stilled n. propyl alcohol was mixed carefully with forty grams
of concentrated sulfuric acid. This mixture was put in a
round bottom flask fitted with thermometer, mercury stirrer,
and dondenser set for distillation. The flask was heated on
an oil bath, keeping the solution at 135° while seventy
game of prepyl alcohol were drOpped slowly on the hot solu-
tion. The distillate was collected surrounded by ice, and
hen treated with potassium hydroxide pellets to remove traces
of acid. It was next distilled over calcium oxide, collecting
that portion boiling; at 85o90° (correct 89°). The n. propyl
ether was finally allowed to stand over sodium to remove the
last traces of moisture.

1:). Preparation of di-isOprOpyl other. .

Iso prepyl alcohol was redistilled, B.P. 80-820.

The same proportions of alcohol and sulfuric acid were used
as in the preparation of the n. prepyl other. The mixture
was kept at 120° while the alcohol was being dropped into the
acid mixture. During the reaction there was considerable 802
gas given off. The distillate was dissolved in concentrated
sulfuric acid, and then separated by the addition of water.
This removed any unsaturated compounds that may have been
formed in the reaction. The other was then dried and neutra-
lized with anhydrous sodium carbonate. A few pieces of
sodium were added and it was allowed to stand over night
before distilling over sodium to remove all the moisture.
s.r. 65-70 {correct Beilstein 68.5.69”).

 

a - ‘1 ...'xull|fl1.1~l|kfl'wnfl~dril~o.v .1. w...) .y, “W.hb A , ‘
, . , 4;

111.1

16.

0. Preparation of di n. butyl other.

This other was prepared by Williamson's method.
Fortyasix grams of freshly cut sodium were dissolved in 160 .
grams of absolute n. butyl alcohol. It was necessary to use
heat and stirring to complete the reaction. The sodium
alcoholate solution was refluxed with 270 grams of redi-~
stilled.n. butyl bromide, B.P. 100°, for three hours.. The
mixture was then distilled, he fraction boiling from.135~
140° being collected. The other was redistilled over sodium,
before being used, to insure an anhydrous product. (B.P.
140.5o Beilstein).

The proof of the structures of these alhyl others
lies in the method of their preparation. Both.mcthods used
are typical for the preparation of others.

4. Pyrolysis of B-alkyl~caffeine others with alkyl others.
The procedure followed for all these reactions was
the same as was used by Allen in his report on.work done in
this 1aboratory‘12).
The carefully weighed sample of anhydrous caffeine
ether was put in,a bomb tube with an accurately measured
quantity of the alkyl ether. The bomb tube was sealed, put
in the Carina furnace, and.heated for the prescribed time at
a certain temperature. When the tube had cooled, it was
Opened, the contents washed out with hot alcohol, and evap-
orated to a dark, tarry mass. This product was dissolved in
hot water, and boiled with animal charcoal to clarify it.
The filtrate was evaporated down in the attempt to obtain

~rr’. 7:

. .1 no... a

J

i

1'7.

crystals, but in each attez-npt crystallisation was impossible.
0n evaporating off all the water there remained a tarry,
brown decomposition product which would not crystallize from
any common solvent, or at temperatures below 0°C. Attempts
were made to bring about crystallization from absolute, 95
per cent and dilute alcohol, carbon disulfide, chloroform,
carbon tetrachloride, acetone, petroleum ether, di-ethyl
other, benzene, and xylene.

We were forced to conclude that it was impossible to
prepare the B-alkyl-caffeines from alkyl-caffeine others and
alkyl others by a pyrolysis reaction. Low temperatures gave
the unchanged caffeine other, while high temperatures resulted
in decomposition of the others to a tarry, brown mass of no
definite boiling point. The unchanged ether was successfully
recrystallized from 95 per cent alcohol by adding a very small
amount of water to it. Its identity was proved by melting
points and by determination of the nitrogen content.

 

 

 

18.

1. N. prepyloxy-caffeine with n. propyl ether.

 

 

gm. 4 W 1 2: I _ 3- 4
Grams caffeine other 2 2 2 2
Co's. alkyl ether 2 5 2 2
Temperature 250 300 300 525
Hours 8 5 8 8
No. of trials 2 2 4 1
Product Caffeine other Decomposed
Grams Product 1.2

2. IBOprOpyloxy-caffeine with n. prepyl other.
Grams caffeine ether 2
00's. alkyl ether 2
Temperature 300
HOurs 8
No. of trials 2
Product Decomposed

3. Propyloxy-caffeine with isopropyl ether.
Grams caffeine ether 2 2
00's. alkyl ether 2 - 2
Temperature 250 300
Hours 8 8
NO. Of trials 2 2
Product Decompcsed Decomposed

4. Prepyloxyacaffeine with n. butyl ether
Grams caffeine ether 2 2
00's. alkyl ether 2 2
Temperature 500 300
Hours 4 8
NO. or trials 1 1

Product Deccmposed Decomposcd

19.

B. Preparation of two new 8-ethers of caffeine.
1. BaPhenyl propylcxy-caffeine. 017H2003N4.

After the manner of Huston and.Allen(1), these
caffeine others were prepared from. the reaction of the
sodium.alooholate on B-chloro«caffeine.

1.51 Grams of bird shot sodium.were quickly weighed
out and drapped into 90 grams of redistilled phenyl prepyl
alcohol, em. lie-120° at 12 mm. This was heated with
stirring and refluxing on an oil bath.until all the sodium
had reacted with the alcohol. To the sodium alcoholate
solution were added 15 grams of anhydrous 8-chloro-caffeine.
Stirring was continued while the mixture was heated for five
hours at 140° on an oil bath,

At the end of this period the alcohol was distilled
off under reduced pressure to avoid excessive heating with
consequent decomposition of the product, The residue was
dieselted in alcohol, and filtered.hot torremnve the salt.
On cooling the phenyl propylexy-caffeine separated out as
fine needles. The product was recrystallized from.dilute
alcohol, the filtrate being concentrated to bring about a
maximum.yield.

The compound consisted of very small, needle-like
crystals which were practically insoluble in cold.water and
cold alcohol. In hot'sater it melted to oily drape that were

comparatively insoluble. The product melted at llOulllo and
sublimed at higher temperatures. It gave a pronounced mur-

exide test, with a‘bright vermildnn color.

20

EXPERII BETA» " Y 1"" ifs OI. PZZJI l‘l'L-PROPYLOX vCAFFEII‘I‘E

Phenyl prepyl alcohol, grams

Sodium, game

8~Chloro~caffe inc , grams

Temperature heated
Hours heated
Recrystallised from

'Yield - actual
" - theoretical

“ - percentage

Analysis:
Carbon
Hydrogen

Nitrogen

Solubility:
Water
Alcohol
CH013

CC14

Determined
62.02
6.15
16.85

Cold
Pract. Ins.

Very slight

Very sol.
Slightly

-

528.08
90.0

1.15
15.0

140°
5

Dilute alcohol

14.
24.3
57.5

Calculated
62.18
6.13
17.07

Hot
fielts to oily drops
9.2 gr./ioo mls.
Very soluble
Moderately

 

 

 

 

JV]
. .. Iwuu‘x. ‘
flu”.- ‘I .

».-.\.P, .

 

21.

2. - .8-Phenyl-butyloxy-caffeine - 013H2203H4.

It was necessary to prepare the phenyl butyl alcohol
to be used in this synthesis. This was done by'making
phenyl propyl bromide from.phenyl prepyl alcohol, forming
the Grignard, reacting the Grignard compound with dry
formaldehyde gas, and finally hydralyzing the addition
compound to the desired alcohol. The exact procedure is as
follows.

An equivalent amount of phosphorus tribromide, 185
grams, was slowly drOpped into 265 grams of phenyl prepyl
alcohol being stirred in a cooled flask protected from
moisture. It was then washed with water, dried'with anhy-
drous sodium carbonate, and finally distilled at 16 mmm,

0(13) . The

collecting the fraction boiling between 120¢125
yield amounted to 185 grams, or about 62 per cent of
theoretical.

A Grignard reagent was prepared from.the 185 are. of
the halide after the manner of deitmore(l4). The halide was
diluted with six times its volume of anhydrous ether. A
crystal of iodine was heated with 25 grams of magnesium
metal dried in a desiccator. Ten min. of the other halide
solution'were added direct to start reaction. Then the solu-
tion was added one drOp a second without cooling, the
completed Grignard being allowed to stand over night.

The flask was then fitted with as wide a delivery

tube as possible, and through.this tube into the Grignard

 

‘ Iaulllll". 1,6»;3Irrp! :3: J .,.r. w... I 9.1!. 4...

 

O
“0

solution was passed formaldehyde gas generated by heating
anhydrous paraformnldehyde(15). The solution was stirred

cry rapidly and was cooled with.a freezing mixture. The
reaction was considered.to have reach completion when a
clear other layer formed on top of the mixture when stirring
was stepped. This took about thirtybfive minutes.

The mixture was then.hydrolized with ice and 301,
the other layer separated, and aqueous part extracted.with
more other. This other solution was dried w th calcium
chloride and then the other distilled off on a steam bath.

Finally the phonyl butyl alcohol was distilled off
at reduced.presouro. B.P. 158—142“ at 14 mm. (Correct ~
Beilstein 1400 at 14 mm.). The yield amounted to 65 grams
or 22.4 per cent theoretical.

To prepare the phenylobutylwoaffeine-ether, 85 grams
of phenyi butyl alcohol were heated to 150° with 1.51 grams
of bird shot sodium with stirring. When the sodium had _
completely reacted, 15 grams of anhydrous Buchloro-caffeine
were added. The mixture was refluxed with stirring for 5
hours at 150°. The alcohol was then distilled off at reductd
pressure. The remaining solid was dissolved in hot 95 per
cent alcohol, filtered to remove the salt, and finally cooled
to crystallize, the filtrate being evaporated down to complete

the crystallization. By redissolving the products twice in
alcohol, each time adding water to cause crystallization, a

very pure, white, crystalline compound‘was obtained. The

 

:FJ‘IIIJB' .4 ,'va‘4llvazF. .«

 

 

25.

crystals were dried at 80°, and finally in a desiccator over
sulfuric acid.

8-Phenyl~butyloxy-caffeine melted at 132-135°. It
sublimed fairly easily above that temperature. The fluffy
white crystals were comparatively insoluble in cold or hot
water. The compound gave more of a red colored.murexide

test, than did the phenyl prepyloxy-caffeine.

EXPI¥IRI.==EENTAL DATA ON PIESNYL-BUTYLOXY*CAFFEINE

CIBHB2°5K4 LI .3?! . (08.10 . ) 542 .0.
Sodium, grams I 1.15
Phenyl butyl alcohol, grams 85
BeChloro~caffeine, grams 15
Temperature heated 130°
Hours he ated 5
Recrystallized from. Dilute alcohol
Yield - actual 16

“ - theoretical 26.1

" - percentage 61.2
Melting point lee-153°
Analysis: Determined - Calculated
Carbon 63.9 65.16
HydrOgen 6.69 6.42

Nitrogen . 16.29 16.33

 

.(‘L'Il Ci.- l.!.'l‘. , iv .
Balanuérv ... ,.J .. u ..:I .. A

 

«24 .

13 BRILHSNTAL DATA (Continued)

Solubility: Cold Hot

Water Very slightly Slightly

Alcohol Very " 7.13 gr./100 mls.
0361:, Very " Very soluble
c014 Slightly sol. Very '

3. Attempted format ion of phenyl-amyloxy-caffeine.
It was attempted to prepare this compound by the same

j procedure wed in the preparation of phenyl butyloxy-caffeine.

However we only had 65 grams of phenyl butyl alcohol with
which to prepare the phenyl amyl alcohol needed. This was
recovered from the preceding preparation.

Phenyl butyl bromide was prepared from this alcohol,
and the Grignard reagent from it was reacted with dry formal-
dehyde gas. 0n hydrolysis there was obtained 8.6 yrs: 13- of
phenyl amyl alcohol - B.P. 150~160° at 12 m. (Correct 8.9.
Beilstein 155° at 12 mm. .

. 1.12 Grams of sodium were. dissolved in the alcohol
which was in solution with 100 grams of anhydrous xylene. To
the alcoholate solution were added 11.5 grams of anhydrous
B-ohloro-caffeine. This mixture was heated on an oil bath at
the boiling point of the solvent for 5 hours. Care was used
to exclude moisture throughout the entire procedure. The
solvent was evaporated under reduced pressure and the residue
dissolved in alcohol. Upon recrystallization from dilute
alcohol, there was obtained 5.2 grams of a white crystalline

 

VWKHJIDI . 3...

Jasmin U3?! }. . . I

 

 

25..

compound, melting at 185«187°. 0n analysis it yielded

23.42 per cent nitrogen. It was found to contain chlorine.
From these results it was concluded that 5.2 grams

of 8~chloro~caffeine had been.recovered.unchanged. Evidently

the boiling point of the xylene was not high enough to bring

about the desired reaction.

EXP 331212.113}! TILL DATA

Materials in grams:

Phenyl amyl alcohol 8.6
Xylene solvent 100‘
Sodium 21. . 18
8-Chloro~caffeine 11.5
Recovered 8~ahloro~caffeine 5.2
Melting point . determined lBSelBBo
" " . correct 188°
Analysis - nitrogen 25.42%
Calculated - " 24.05%

4. Conversion of phenyl propyloxybcaffeine and phenyl
butyloxy-caffeine to hydroxywcaffeine.

One gem of each of these two new caffeine other: In
heated on a steam bath with 10 mls. of 10 per cent hydro.
chloric acid. In each case the ether dissolved, and there
was formed a small amount of what was assumed by the odor to
be the phenyl propyl and phenyl butyl chloride. The volume

of liquid was concentrated to about four mls. and on cooling

 

legit-'1‘ lentil-Ia v1 trul- IMF LE. ~3le 1d

 

26.

there separated out a batch of white fluffy needle-like
crystals. 0n recrystallizing from.hot water, there was
obtained a good yield of hydroxy-caffeine, melting to a dark
brown liquid at 550~540°. The proof of structure lies in the

method of preparation and in the physical preperties.

EXP ERIE-TEETAL DATA

Materials used Phonyl prepyloxy- Phenyl butyloxy-
Caffeine compound - grams 1 1
10 Per cent H01 - mls. 10 10
Time heated on steam bath 20 min. 35 min.
Hydroxy—caffeinc . grams .58. .62
EePe "' determined ‘ 330‘340
" correct 345

5. The attempted molecular rearrangement of phenyl
propyloxybcaffeine and.phenyl butyloxy-caffeine.

It was attempted to rearrange these others by heating
in span and closed tubes. After heating the contents of the
tubes were washed out with hot alcohol. On.ewpcration to a
smaller volume there was obtained in some cases a crOp of
crystals, each tine proving to be the unchanged caffeine~
other. In other cases, where the temperature had been higher,
there was more decomposition and discoloration, resulting in
the recovery of loss of the caffeine ether. The structure of

the recovered product was proven to be the other by its phys-

ical preperties, and by the fact that it formed.hydroxy~

caffeine on treatment with.hydrochloric acid.

EXPERIHENTAL DdTA 0N ATTEfiPTED REARRAHGEEEfi

1. Phenyl-prOpyloxyocaffeine

 

 

Trials 1 e ‘38 4
Ether used - grams .5 .5 .5 .5
Temperature 150 225 250 325
Hours heated. '8 8 8 8
Open or closed tube Open open closed closed
Remover ether, grams .35 .19 .11 decomposed
2. Phenylwbutyloxybcaffeine _ .

Ether used 1.0 .5 1.0 .5
Temperature 150 225 250 ' 325
Hours heated 8 8 8 8
Open or closed tube Open. Open closed closed

Recovered other, grams .75 .12 .16 decomposed

II: It: IIAL .1 Dana

 

 

28.

SUMMARY AND DISCUSSION

The first purpose of this thesis has been to discuss the
methods thereby it was attempted to prepare the B—alkyl-
caffeines. Ehis was done by the pyrolytic reaction between
8~caffeine others and three of the alkyl others.

It has been shown by Hurd(1), that an alkyl other will
yield on pyrolysis, an aldehyde. It has been eXplained.that
the probable reaction between the aldehyde and the caffeine
other is as follows:

Caffeine « OGEHV + 0 3 g — CgHs 300° 9
Caffeine ~0235 + 03370003

An examination of the equation will BhOW’that the re-
action_involvest

l. The formation of a caffeine radical and a propylozy
radical.

2. The breaking down of the aldehyde is as follows:

H H
02H5 - C n O to 6235- and -C u 0

3. Finally the 8-alkylwcaffeine is formed. But there
is involved a migration of hydrogen from.carbon to oxygen, and
a migration of oxygen from carbon to carbon before the acid
can.be formed.

The very complicated mechanism of the whole reaction is

29.

probably the reason for failure here in preparing any of the
8-alkyl caffeincs by this method. It would be interesting
to heat propyloxyecaffeine with.propionaldehyde, as a check
on the above explanation.

It must be concluded then, that it is impossible to pre-
pare the 8-alkylecaffeincs by the pyrolysis of the 8 others
of caffeine with alkyl others.

Secondly, two new caffeine others have been.prepared from
the action of a sodium.alcoholate on Buchloro-caffeinet

l. 8-Phenyl prepyloxyocaffeine, a white, fine, needle-
like crystalline compound, melting at llO-lllo, practically
insoluble in cold water, and melting to insoluble oily drop-
lets in boiling water.

2. 8-Phenyl butyloxy-caffeine, a white, fluffy,
crystalline compound when crystallized from dilute alcohol,
melting at 132-1330 and very slightly soluble in cold and hot
water.

Both.phenyl prepyl and.phenyl butyloxy-caffeine when
heated in an excess of 10 per cent hydrochloric acid, will
form hydroxy-ucaffeine. The moxide test is given by each of
those others.

It is interesting to note that up to phenyl propyloxy-
caffeine the melting point of the ether is lowered by the
introduction of a methylene group into the chain. Phenyl

butyloxyacaffeine however has a higher melting point.

30.

Compound E.P.
Phenyl benzyloxy~caffeine 172-175.5°
Phenyl ethoxy-caffeino 142-144 .4.°
Phenyl prepyloxy-caffeine 1104.111O
Phenyl butyloxy-caffeine 132-1330

A possible explanation for this might be that up to and
including phenyl propyloxyvcaffeine, the aliphatic
characteristic is predominant, but beyond this point the
other character becomes more evident. I

Finally, it was attempted to bring about the molecular
rearrangement of the others to the 9-alkyl-uric acids. This
was done in both cpen and closed tubes at varying high
temperatures over different periods of time. However either
some of the unchanged ether was recovered or else there was

decomposition.

 

1.

2.
5.
4.

5.

6.
7.
8.

9.

10.
ll.
12.

15.
14.
15.

BIBL IO CRAP HY

E. Fischer, Ber. 22,1785 (1884), lbld.§2 569 (1897)
Ann gig 261 (1882).

W. Wislecenus and H. Korber, Ber..§§,1992 (1902)

H. Bilts and h. Bergius. Ann‘ggg 59 (1917)

A. Baumann. Arch Pharm. Inst. Univ. Berlin lg, 127-147
(1913).

R. C. Huston and E. F. Allen. J. Am. Chem. Soc.lgg
1556-1359 (1954).

Ann gig 268 (1882)

Blltz and Strufe. Annuéié 199-200 (1917)
Hard. Pyrolysis of Carbon Compounds. honogranh Series
pp. 198-199. ’
flinchelivood and Askey. Proc. Roy. Soc. (London) gigg
215 (1927).

Fischer and Reese. Ann‘ggl, 356 (1833)

Norton and Prescott. Am. Chen. J. g 241 (1884)

Winston F. Allen. Thesis for Degree of Ph.D. from.this
laboratory, pp. 79-80.

Soc. Chinie. ‘23 1410-17 (1931)

F. C. Whitmore. J. Ame Chem. Soc.‘§§ 1561 (1933)

see p... a la Soc. Chim. pg 682 (1932).

MEFT—l-‘fifi'fl‘l‘ '-'

 

 

 

III. III I I‘ll" 1'1. I Ill

. .Ir a. Isaialrvflnlpfii... c.rtuh\7‘.0.h\.‘nl..{ $.n .9} .av. I .I‘.

 

 

I'll I III! III] .|J~

u . . '
-.. .. .h In 0 a. .I. .. \i da. ..-ufl1hfl.lu¢.dl.bll

 

I \llllr a II??? ‘l..I'e
.w e ....I I . l o ‘ z .

 

 

\V l“! is. (Flcf'w' \ I.

‘ .1 n' Its" a v e \InhOI

 

 

 

a.
t I

   

 

"Iiililflljillfilfl’ilifllS