1 THE PREPARATION OF

ANHYDROUS OXALIC ACID

II A NEW METHOD FOR THE
PREPARATION OF FURFURAL

BY

VANDRRVEER VOORHEES

THESIS

FOR THE

DEGREE OF BACHELOR OF SCIENCE

IN

CHEMICAL ENGINEERING

COLLEGE OF LIBERAL ARTS AND SCIENCES

UNIVERSITY OF ILLINOIS

1921
 Digitized by the Internet Archive
in 2016

https://archive.org/details/preparationofanhOOvoor
 .

UNIVERSITY OF ILLINOIS
2

o Jarn^Z
X
<M

THIS IS TO CERTIFY THAT THE THESIS PREPARED UNDER MY SUPERVISION BY

y Ail.aSRYJlSlR , .
VOOMM. S .

ENTITLED, I......T.HE...PREP.AaAIJ.0iI...QP....UmYI)RQU3....01A-LIQ..AC.I13.

...I.I. A...m...MS:/IKQ.I?...£QR.YXPPI...ER:5P.ARATimi..Q.P.„^

IS APPROVED BY ME AS FULFILLING THIS PART OF THE REQUIREMENTS FOR THE

HEAD OF DEPARTMENT OF....

^Kv
 -in

CIOWJJI YTl 2 XaVlMU

"i;

»:

-.)
: « X - • !» ' lA
.
\ »
'-' '

• * "•
' --V
jv
A

***
Mv .

.i7.^,^*,. (i»
^5

va Kor^ivaa^T Jd vw h^viu cft^iiiT ji-rf TAKT t t>i 4

i ii‘; ' ^
'
^')
J
i,-

,.SI.|(LJ|!UA-C.u . S
'

5.1

r*'^ f-,

3HT aOT Pt r UO ljlt>‘i rlW f ji/M I jlY lU''l 2/. 3W Cn

|; ;

—
V. . .
. .
.
* ‘.„ >Tr/ ‘
jJ- ^
„
*

'io
'

3 !uiL>aa^
Mi^
<fii f^- A
«M. 'n o»>>ri4i«^ !>»'• (•••^'.IwiVlf*'^
'
A* /
.

. i
.
.'•
.
?
I
> ..^w»»i m i. »kmf f .*•
V *
"yTr *

-;ariiP
rtt
• fr'^
-
-W«9
,

,
f

^ '
* rt

?
t
r
'
'A
</

V
\

'->
‘ '
- r,f V , \ 1
\f '
)
- * ifc>
* •. » •' 'I

9 S.tMi.> at iofj#t\»-'<<l
*0

. » > i* '

m 7.

«tj"
i 4i

4i
»;r-
'
-»»»m(i*>.^i(«r 30 I KHMfW. ‘’30 nOUAifH
 TAPLE OF CONTENT?
Page

I Anhydrous Oxalic Acid

Introduction 1

Experirr.ental 3

Di scussion 6

Drawings of Apparatus Appendix A.

II Furfural

Introduction 10

Di scussion 12

Experimental 14

Notes 21

Conclusion 22

Summary 23

Bibliography 25

Drawings of Apparatus Appendix B.

 5

70 i.nt;
'
.?

-'
i!2 ' /T-

y-^'>4 f: ' - ^z— '-i-'.'A ,t*^T* •

^ .-a
r<4

X-»T5T’= ^

ac i

r- « *- f‘'**T fr.t-B'
5.ttL'ic^i
n:

AOXSwi«UjtQ
in'

% ii^sayAit ^
3
^'r f^, n^'-j

^4,* ^fi,. ^;

...iJ ^
- *V - *i
^ <>?.
? ^ !^'-;' .' 4^r:f aT€f5:'.|A^‘Jb U '*5^
'^j

Ifc'
 1

INTRODUCTION

Certain organic reactions, such as the preparation of oxalic

esters, require oxalic acid entirely free from water. Since the

pure, crystallized acid contains two molecules of water of crystal-

lization, this must he removed before the acid may be used for

these purposes.

Oxalic acid cannot be satisfactorily dried in an ordinary oven,

since it sublimes arpreciably at temperatures below 100©, and above

(1) (2)
this temperature decomposition ensues. Beilstein mentions that

the water of crystallization may be removed by standing over cone.

HgSO*, but it is obvious that this method would not be applicable

to the preparation of any considerable quantity.

In order to overcome the above difficulties. Dr. H. T. Clarke,

of the Eastman Kodak Co. h&s devised the following process (as yet

unpublished) which may be carried out in any laboratory. Finely

crystallized oxalic acid is boiled with CCI4, the vapor of which

carries out the water according to the law of partial pressures.

The vapors are condensed into a gravity separator where the water

flows off and the CCI4 is returned to the flask. This is continued

until all water has been distilled out, the product being obtained
by filtering off the CCI4, the last of which is removed in a
current of dry air.

(1) Gay Lussac,- Ann. Chera. 1-30,

(3) Handbuch - 1 - 639.
nA

Ll «. rfj

*\ \ 1 ' '
‘
r, * w* *• , V- ii
'
i
>
»!

, "tr . i. V' - - >?, '

r *.• :• s >>'v ,.
« ;i, ,i*' 1 4 ^ 0®^!
-> vn i()t /

'A-

.... n..; . v.-jf?.

4.
•'.• !
.. . -/r'Ts*. :t 1 ^cJ •jfl;; I

n;cj :^c:
•'.
.17 ':V . *(f^' '.'K. ^ ^ itti ^ •
f

,|
. - • - - r .'i”’ .
• f ». 0
, - - I ^ ‘
, •‘i I- • -;.

‘.v. t-
* h

=Sa
1)' r
'
- rr'/. luJ^ ' )

T!?r hi' I if ^ wmnimi

 3

The purpose of this research was to check over the above metho:

and determine the most suitable form of apparatus to use; one

which would give the least loss of CCI4 and dry the acid in the

shortest time. Previous work with this method showed that*^large I

loss of CCI4 could not be avoided, thus making the process an

expensive one. To determine where this loss occurred, a number

of experiments were made with different forms of apparatus, the

loss of CCI4 being observed in each case.

ETPERIMENTAL
Experiment 1. - Two Kg. C. ?. oxalic acid were passed through
a 30-mesh seive and placed in a 5-1. round bottom flask with 5 Kg.

CCI4. The flask was fitted with a glass stir of the vdng type

reaching almost to the bottom and sealed by a column of mercury

about 3 inches deep. Connection was made to the top of a vertical

condenser by a tube 50 inches long and 1/3 inch diameter, wrapped

v;ith asbestos to decrease condensation. The condenser, 30 in. long

and 5/8 in. diameter was connected at the Icv/er end tc the water

separator. In Dr. Clarke's apparatus this separation was effected

by means of a simple trap with a funnel set into it to catch the

liquid from the condenser. The objection to this is the large

surface offered for evaporation, so a modification of this separator

was used in this and follcwing experiments. This device, shown in

Fig. 1, consisted of a glass tube 8 in. long and 7/8 in. Ham. with
a small tube sealed at the side for a water overllow and a second

small tube leading up from the bottom to within 1/4 in. of the water
 \
\jrr^

: ‘.'it .ic t^fO ••;..

’
,‘, OJ .-;
*
'« I

. . ^ t; " J '

»v .^:
. c < 0 L^ '

rori if - ©
•Cl V*. .
'..

I
•

f t‘Tr; ?;r -.v, c .‘'n«

•iVl
'
7
-; ;*i ?
.'-
/
•
. T

.•^-> Vt.* ,3"

• J..
V'Ti-'
• r V-7 it -A

' . >
;
-
rt,
. '.V
A .V'i it'

1 •, * V ,
«
«.
t .

?• )“ - ^t ;
. ' i.'" 0 t. i ••

(; f.
.T X .

'
- V -

6* J
^ h. q:
,

'.t

^ i
4. iff' *i .';- : ::

•
/.
•
. --Cl'i ,v ^ ^ ^ .v» .

“
'i
.,- . ... V .
' .•• :.-y r. i-

-w Lr'.'iS ti . .' - v1
'•"
'
i • < ' ” i’i
. . • .

‘ '' .*; :ir>i >

C
E.rl;’ /
‘4 '

<

•
• ^

t; » iiti ijifi
outlet where it was bent horizontally and connected to the flask.

This maintained the level of CCI4 just below that of the water
outlet, and the water, being much lighter, collected on the surface

until the layer became sufficient to reach the overflow where it

ran out and was collected in a covered beaker.

The connection between the separator and the conienser was

arranged with a capillary open to the air to maintain atmosp'heric

pressure and insure the proper adjustment of the liquid levels.

For this reason also, the return tube was made 'with an enlargement

containing a capillary to prevent siphoning the contents out of

the separator into the flask during operation. The returning CCI4

was delivered by a small tube extending well below the surface of

the mixture in the flask to prevent vapor rising and escaping

through this last mentioned capillary. A small funnel was sealed

within the separator just below the condenser to deliver the 00X4

below the surface and prevent its being floated out with the water
by surface tension.

Corks were used at the condenser connections, but a rubber

stopper was used at the flask, owing to the difficulty of making

a large cork stopper carrying two tubes and a stir, sufficiently

vapor tight. Hovirever, the rubber was abandoned at the end of this

run as it was found impossible to prote ct it from the solvent

action of the tetrachloride, altho tinfoil and collodion laquer

ware both tried. The rubber also showed a tendency to discolor

the product.
 .
‘
:
,’ ,
.'. i’ H ^ . J AJm . .. tXJ

^
' •’’
.
*

i - tl •
t| bttii :
- r, M*r
-*

\,
lit*.., .

'
Q< 0 c:
•

S i .
,
'

1 i. a!
‘

;»:* ..< ,
f.‘Ot
,

.
•••’ ':> ' • WjuMA '

; t
I ,
•.

'- •

. Mr* • ^ . 1 ..
.- «. ‘s. . 'i
,
.
-

•>
J- l;o t- ^ ci
:’
f.-

" V._ .-:

4 *•'' o.r' '• t -

•' -
-S' ^ i *
' >
^
] :> ,: T, t r X - : ; ; . .? r. ^ . "-i q
'
•> <•

“^ 5*
,n f .
*v- .- . t .f 1 . i
'.u t .’
,
-

'

.: . ;. -V .:» -rtjvaiq o* x*' - -

fisc '
•'"•
.
'
.'?o t-.-ll r ! 7 0 liai tiit
* ;:
• ,
?.

• -
' .:- .•:• -U' . V i x^sfe..

c„' .
1 .
-r
'
.;, ..aX ,
• .a 4 ^,
.'* ‘'^
. - r ; A 4 , :,T; 7 rfir

'

.
• »

;:c.-tja:'i;
' '* ‘

.'•
^.. ,.., .< .: •

.. : . .:.
' .' •'
•

'•'•
'; i ^'il •Up.'? r*.^ •> ' fV; 5 *.••;
^ QT ,
’'

•* '-
'• ivi
‘ .
-^'.fv:^:
r

•’ ' ’ •
.:' ••
'- .* - r-I •
c ••

i ’..liv. ,
.

; :
•
i;a.
'
:;, ii' ‘
c r r ‘ .
- A's
.
i

.
V

'
l i
Y-.Tvt'fit i* . lii ii'J
 - 4 ~

Distillation ovsr a direct flame vias carried on for a total

period of 30 hrs. At this time the CCI 4 returning to the flask

had lost part of its milkiness due to entrained water, but had not
become clear. Gome trouble was occasioned by the return tube

becoming closed at the end by an agglomeration of oxalic acid

which required frequent removal. After cooling, the acid was

filtered off by suction until no more CCI 4 could be obtained.

The remaining CCI4 w2i.s expelled in the oven at lOO , but some

sublimation occurred at this temperature, causing the finely

powdered acid to sinter together into a mass of delicate needles.

3900 g. of CCI4 was recovered, denoting a loss of 1100 g.

1350 g. of anhydrous acid was obtained, which is 94.5^ of

the theory. The loss was most likely due to sublimation in the

oven. Titration of the product showed a water content of less

than .01 mol. HgO.

Experiment 3. - In this escperiment several changes vyere made

in the apparatus. The length of the separator was increased from

& in. to 14 in., and the diameter from 7/8 in. to 1-3/8 in. This
was done to allow the water more time to separate before the CCI4

was returned to the flask. The tube returning the GCI4 was ex-
tended just through the stopper in the flask, and a loop was made

in it to prevent the reverse passage of vapor. In this way,

plugging of the end of the return tube within the flask was effect-

ually prevented, A cork stopper was used in the flask and painted

with sodium silicate to render it impervious to vapor. The

 -

t >

nii .tjO'i ^ - .^ i- 5>,^

j
Kci 1 .-1 i

r,%tf • ? • V.; «>,•• ICTCCX

<i'

V •
> ^ •
*
c' tiKH *T

*«*

.-•ft

..Xl'’ - s.« •• »/ . 'I,

.-
.;fn S
i|." ^

Ufcrtx» 'toJ6« *Xt®' - '-5 I <

b .
i '!t *. O'i’j t ..0,t i.'/ .'
•

V
•1 -'. •
.
5’« -T,- ^ . .,

' f > '- 1' •

•
. 1 ^
'

:V .
L f
«

l—J
il ( '-'{.' -i" ' ^ -^j ^_ A.' ; -- ,,, .v.oerf'j

V
ic%

'
’.«
i:

'

.,»- ‘'A'* •

’( * f
w *

k»*»» W T- % .^ . juT A.<::?<wp >> V .

IM»’ A IC . *
t-r ts: ^ aj 4- *•/> X.V^ Awl A

;
^
!:
'

? 0 X>
^ .. »t X' ;ti £ii^ .
'
r' .
;

' <.•
O .
fill'*.,

j : : .

- < f <- . / :. ; .. / rii3 , 'a n

iir.^ :
" : y (..‘j: - . ‘-ir,<- .
'
 - 5 -

capillaries previously mentioned as essential to the working of

the separator, were replaced by connections to a small vertical

condenser as sho'^n in the sketch to prevent any escape of CCI 4

vapor. The small funnel which was selaed vdthin the previous

separator, was merely rested on a constriction of the wall in

this one. This was done to siiiplify construction. The entire

apparatus is shown in F'ig. 2 , and was used in succeeding experi-

ments. A water-bath entirely surrounding the flask was used in

this experiment in place of direct heat.

2 Kg, oxalic acid graded as before and 5 Kg. CCI4 were placed

in the flask. Distillation was carried on for a total period of

18-1 /2 hrs, 2500 g. of CCI 4 was recovered by filtration from the

anhydrous acid, showing a loss of 2500 g. Part of this loss was

due to clogging of the return tube at starting and overflow of

the separator. The product 'Afas freed ffom CCI4 by spreading on a

paper placed above a radiator for a few hours.

1450 g. of anhydrous aoid va,s obtained which is a trifle more

than the theory (l430 g.). Allowing for .02 mol of water, deter-
mined by titration as before, this is practically a quantitative
yield.

The use of the ;vater bath effectually prevented the formation

of a crust of acid about the upper part of the flask, this having

been previously observed by Dr. Clarke.

Experiment 3 . - The apparatus was used just as in the pre-

ceeding experiment with the exception of the water bath. It was

felt that the slight advantage of the crust prevention was more
 ;a^:ov :
-U » .

V U-^V .... r i:,. -r.- . . .

10 -.^ , T4 ^ in V

r..!%;' j

til

‘t.’ L^-<7 '

''
:/T^:
„ 'i;V .VI
•
M. ^

*
'

; X o? t- 'lo ^

- .•<

r ’

.‘ y-'V
•
- s

x.-^ i-
.- ^ ••.S','h>’ -c '_.n I*,.; ^‘>0 O' ^v*
V
.:'( ^v- u:i.
•fA- 'i5
»'

. r .
,• > - -

t"‘ t
;.T , X t ... .
r U t .A'

o’ i.'.cr ;. . .
... :) < . ; -

ill
•
\ ,i. t
,
: . V

>."

^ IV4
•.

•,c..-f!Mf'
‘'- *
7^.; • .
•
(U «•» I

f‘

t
•
-/*.
( ^
* . -4 -
m
h
*"•11 0 ^ r% it

' * > :
:X.\’-v <- :
V ia,- j

. * .. . V 4

4
^.r n ,?..
,4, -X

M *'TO«il 9 .^1 ’’ ?«

r.t^i^ux-virvf^ ^Ci^' *) »/i -cf- V c_ v'j

. 4 *-

'.*

i * Jtmk i k JU
 - 6 »

than offset by the inconvenience of handling such a cumbersome

water bath and continually supplying v^ater to make up for that

lost by evaporation. All connections were wrapped with strips

of cloth which were wired on and painted heavily with sodium

silicate. Very good seals were obtained in this way. The extracted

water was collected in a covered beaker and measured at intervals.

Distillation was carried on for a total period of 16-1/3 hrs.

Water was collected as follows;

After 7 hrs. - 360 cc.

” 13 n - 465 "

w 18“l/3" “ 545 «

The theory requres 570 cc., the discrepancy being undoubtedly

due to evaporation from the beaker during operation.

The product was dried on the radiator as before and the theory

yield was obtained (1430 g.) . Titration showed the presence of

.03 mol. water.

Loss of CCI4 was not determined in this run because of an

accident which occurred during operation.

Experiment 4. ~ This run was made largely for the purpose

of checking the results of the previous run and determining the

loss of CCI 4 . In order to minimize back pressure on the flask,

a steam jacket was placed on the vapor tube leading to the con-

denser as shown in the sketch. This was advised by Dr. Clarke

and entirely prevented condensation in this tube. Apparently
the previous losses of CCI 4 were principally due to loss of vapor

at the mouth of the flask and around the mercury seal of the stir,
 j.,

\
'/ 1

.i

:i:,!u n.I •
^ - M-*’’

,, l;'-> '
r. -dr hcsr&tM hfi;^ ^pt ^c.4^-sW'' '‘-'-5'

M
 . 7 -

and consequently any increase in pressure would increase these vapoi

losses proportionately* Greater care was taken to inspire a tight

joint at the flask, the stopper being painted with sodium silicate

both inside and out and the entire joint wrapped with cloth,

securely wired and painted with silicate solution.

Distillation vvas carried out for a total period of 23-1/2 hrs.

during which time 555 cc. of vvater was collected as follows:

After 4 -I /2 hrs. - 120 c.c.

n 13 n » 375 "

w ie-l/2” ~ 520 «

« 23-1/3” “ 555 «

A decrease in volume in the flask during operation, of about 500 c.c.

results from the removal of ’water. The anhydrous acid was filtered

off and as much CCI4 removed as possible by pressing and suction.

3950 g. CCI 4 was recovered. The product, which was in a

damp condition, was weighed before drying to determine the loss

of CCI4. It was then dried on the radiator as before for a period

of 3 hrs. after which it was bottled and weighed.
Yield - 1420 g., practically theoretical.

Wt . before drying - 2350 g.

” after « - 1430

Loss of CC14 due to drying 930 g.

930 3 9 50 « 4880 g. or almost 5 Kg., the amount used.

Titration showed the presence of .03 mol of water. Indications

are that some of this, at least, was absorbed while expelling the
Xf"
 « 6 -

last of the CCI 4 on the radiator, the humidity being fairly high,

at the time, and a slight crust being formed oer the surface of

the acid after exposure. It is to be noted also that the product

from the first run, which was placed in the oven to expel’^CCl 4 ,

contained the least water (.01 mol.)

DI SOU?? ION

The success of Dr. Clarke's method for dehydrating oxalic

acid is quite evidently dependent on the economy of carbon tetra-

chloride consunp tion, for it ^ves unquestionably a pure, dry

product. I believe my exr>eriments have shown ;jU3t where the losses

of CCI4 occurr, and the manner and extent to which they may be

p re vented.

The apparatus which I employed is practically 100*^ efficient

as shown in the last experiment in regard to CCI4 retention. The

loss of CCI4 on drying of the product is, in so far as I can see,

unavoidable on a laboratory scale. This loss could, however, be

greatly minimized if some form of pressing machine were available.

When using the laboratory type of suction filter, I think one may

figure on a loss of CCI4 equal to 50 "^ of the weight of the crystal-

lized acid used.

The most undesirable feature of the apparatus is the impos-

sibility of using rubber stopper connections. Sodium silicate,

as I used it, adheres to the flasks and tubes, dr^'ing in horrey

lumps which can only be removed by prolonged soaking in water.

In one experiment the stepper was so securely cemented to the

r iiL
'

; J

-
6 . i > 9A
i

'^,J'C;. to'. V . r»" f ,. : - •

»;' •
-
:.: -.-.. -6 o,.-. 'a: zr.

,r

/ % Cl '
. • ,;J
>’
v ^ .

.'
"-.
/«'.• .T“G« e/f,' r:' i ,.Vs , ,. !

'•
, M. r.L' cev Xf •

tc ;rct-' :- iat

<;,^ ., pfiJ jvVcii •

' v«.-T cu’ne-.' .

A'’ •

5.

ias / V ; ^ I J^err- ^ .

J i .i
:
y' ’’i".

' •

,-: *:y
1
/ . 4 '.
o*:-.; :
; ;

'T 1^
” f

*
“ e* V'ip: I t
.. .

^ 4. ^ M / c.

" p4-T
a ., r: . U. ^ .
-
’"T «2 %- .
1

•^t .* - V’ 'tc '’\*J Xv^ f *

^>' 4 *

'j
1 ;.- Ji

{{->.' . a .
i‘

•^ te' .»

i.- CM- . i

.4

,j.^
 - 9 -

flask that it had to be cut and removed in pieces. I see no

solution to this difficulty except in the construction of special

flasks and connections of glass arranged for ground joints or

mercury seals.

It has been suggested by Dr. Clarke that the use of calcim

chloride to dry the returning CCI 4 toward the end of the reaction

mig^t considerable shorten the time of operation by removing the

entrained water. This was considered but abandoned because of the

manipulation and loss of CCI 4 which it would entail, esp^e daily

as the drying agent could only be of use during the last 3 or 3

hrs. and I felt that this saving in time would be insufficient,

particularly as the apparatus which I employed required practically

no attention while running.

I might add that when the reaction is properly controlled,

the second condenser is unnecessary and that the capillary arrange-

ment described in the first experiment is quite satisfactory.

 • *'. v.H * -h:/

>

j '*
*- u -
».
p *
*
I
. : ,
• ’

.}" ,
’VI •!! ^ '.t .vbb™-, ”s *< V' ft'
.,p- ,

(i't}i\.p.m'X .'
4 eg qr «i «?!<>•; bac 4-^ ^
'

av '

‘
, A7 .'"1

to 0i>i7tiln^taq^ 94^ #;;eo>5W yill/iiiftit’ «id^"o4.,«Oi:i4/Xee

^ ST' 'M
t:t
• '.^
* Ji. y

^ 10 ^re4J^a Ig |j

.-Q _
iJ;
ff:^' y;;. •>.
,tt r i.

mrjLe-JUjg 10 4{-« ><rd* o.*?tji*ft? . .^ rd te/e -

^tj

. -ooitfCP-^ *10 Oi4.%o) ^4;r x^t b7 [-

6>f.? aptvowtsi Y,d tfoi^eioqp? Id diets' vt»t ctt^5St'e%

r ’
•-
.
•
.’.** ’ •
’« ‘*v B*'v.
."’

j
orfi. to ®aiV«o4cf ^oroOiwd^ tfwd
'
#“
•i‘“ 1.^' V
.V 5 ''
‘^
''

P‘
[!' i.;i' ., .
'-
,

^ „;"0
Yf :Aigg.-jatf% f Iflifixe t/, il5*r
•‘-
4L-»«
••il** '-^1
’V t 2»- :-M
f, « TD .j^ Sa 94 9Ji^ »ajt*i/S> 9a a ^ ad xXa.^
’«’T
to,
-

1^ V •

'> .• ... k.
SiS. *if 4 f iOittgsai ed Um
Jt,Xi^«» ‘iA'rtii Btit
•

^:xHi.i ,f.tdf
.

- ^
‘

‘' ‘
!0 fl ^ ll '

•-
^ipi^t^CJirq b6*x Ufpm h T f:s)/^A%-^:qK tXi4imUi0^ feiJ

-."
**
at*'.
If'

-
'

j
'

i
tpiXgtdaoo ;Y.Xtec5piq' X r-g fcj 0*: tx ^
*t- ^ Im
’
''’"

?cisX(K T
', 1,

»5 i}*r3 J *r^ ^•**W to* « i\i;rv,a!>in>iia to,40 «

4 i

"V* - - •
;
’*'?
'

*B ^
'

“' **' -'

'
Yto?.^teX^J5 i e;fiic sX .tiit v--l? -^4 J* f. •
,

'^ '

-'“jj'
'

r- » '. _ I' •Sti^ :' 4J I i e *::ii

.
,^. ^.- V.r'
‘^ ’'--•1
•

', / ,- ltf ., k<^^

6'ii,
 A NEW ItETHOD
for the

P REP RATION OF FURFURAL

j9
 10

INTPODUCTTON

Furfural (furfuraldehyle, furallehyde, furfurole) when pure

is a colorless oil, slightly heavier than water 'with a boiling

point of 161® (Esilstein) . It is a cyclic ether aldehyde having

the following formula with all the

properties of an aldehyde and many of the properties of the aro-

matic series. It is soluble in v;ater to the extent of 1 part in

11 at 13® , (Beilstein) and considerably more than this at higher

(l)
temperatures. It is also quite volatile 'with steam. Cohen

gives its formation from pentose by the loss of 3 molecules of

water.

Occurrence - A glance in the literature will show that

furfural occurrs widely scattered among the processes of annlied

( 3 )

chemistry. We find it in vdne, beer and fuseloil and in waste

(3)
products from sulfite pulp manufacture. It may be obtained by
(4) (5) (6) (7)
acid hydrolysis of sugar ,
bran, lignin, cellulose, and oxy-
(S) C9)
cellulose and also by hydrolysis of bran with zinc chloride.
It results fromi the acid hydrolysis of pentoses (arabinose and
( 10 )
xylose), and pentosans , this reaction being employed in the

laboratory for the quantitative estimation of these carbohydrates.

In brief, it seems to result from the acid hydrolysis of a great

varietv of vegetable products.

Preparation - A great many processes have been devised for

the production of furfural, a number of which have been patented..

Almost vvithout exception they consist of some form of hydrolysis,

 - 11 -
( 11 )

this being accomplished in some cases by steam under pressure

or by various strengths and kinds of acids.

The laboratory method which has been in general use is

( 12 )
essentially that given by Eeilstein as follows; bran is di s-

tilled with 30“^ KgSO^, the distillate containing less than 1*^ of

the oil in solution. This distillate is concentrated by saturation

with salt and redistillation, half the volume being collected.

The first of the distillate is most concentrated and the oil will

separate at first and later relissolve. In practice it is rem.oved

as soon as no more is seen to separate from the water. After each

distillation, the water solution is saturated with salt and re-

distilled as before, this procedure being continued until all has

been obtained as the oil.

(13)
Stone found in 1691 that a very much larger yield of fur-

fural could be obtained from corn cobs than from most other veg-
(14)
etable substances. Following this, Noyes ,
gave directions for

its preparation from cobs using constant boiling (SO^) HCl as

hydrolytic agent. During 1917-16, considerable quantities were

made by the latter method at the University (Dept. Chem. Mfrs.)

Results from this work show that an average yield of 10-12*^ of the

weight of air dry cobs was obtained in this 'way. Also, for each

150 g, of oil, approximately 30 1 . of solution had to be distilled,

which required an average of 15 hrs.

Use - At present the demand for furfural is rather a limited

one. It is used as a reagent for the detection of aromatic primary

amines and has been suggested as dyestuff intermediate. Colors

 -i* '
..-‘JTw,^

Sf, .

'
\>i3

« '
I' \ .?
L

r -1 • ,
• t
’
i,
'
<

««*A
' ^ ...f 'iir’ti'^ \
J. jJ .i >fc»

'
'’^ : a
- ,
‘
V.
T.

.c X

A" r, i/O^ ;. i/ K
X.^ : .
4

, ii

‘
3 .
'• '
"T !
-t- -

. f ; ^»r: iJ vxt\^o W I.
J . , ) V j' i. 'M'

r:> ^

r. r.-
:. J’

''
t
•
; '.X n T .: ,'i U’4(^.'- C"'.
I

r^f
'
n. I : .
;^ii> rXiX||<l l,OJ/a. |;

rf
.-.s^ '
< '
.,
'

i •§?( .'I
- '
•‘V #
^u\ .'i
/ .
.» •
•• • "jri/Tft TO > s?i Liti

» *-
'*
i .
*p,‘rr ‘
P '

^i
'
r
^ * 1. * ' ‘
-t *»• t>« A
-jt.v t -r* i 1

/
•W-
! .

‘ - :y -t-:.%-: i-

••’t
;

- r - - i .;
.-a* - •

,jT

i:c;-' ^ 1 ‘
^ 1^. &; '
j|
•;
'

p ^ c
. :.. i
f.;.-- .\'rci<r >* *rf"

• 'it 0 :• •; :ii3€^X^P :.i4' V * *

.

4r.ixgTq“ . ;
, n ~ 4-
lA c. ‘i;-'." 1 J
(

- *i
•
^ /5 ^ - *1 , :

n ...

itr. t . .1 3 c-.' ../

^ * V ' ta t 'Tl(«'
n'
••' ’^TT" -j,4- yg, <er jg^ j-iwWugM Mw» —
'
•
 - 12

made from it, however, have proven fugitive to light* Some use

hae been made of it in perfumery and as a vermicide.

DI SOT SION

As far as I am aware, no method has yet been published by

which furfural can be obtained from vegetable matter without the

distillation of disproportionately great voliimes of water (or acid)

and the subsequent concentration of the distillate. This pro-

cedure is necessitated by the fact that the distillation must be

extended over a long period of hours, since the rate of hydrolysis

is only moderate at 100©. The distillation m^ust also be maintained

quite rapid thruout this time in order to get a good yield, and

consequently a large volume of distillate collects and must be

concentrated to obtain the furfural in solution. Several attempts

which I made to complete the hydrolysis previous to distillation

and then distill off the furfural 'which should have been formed,

proved failures, very small yields being obtained. This led me to

believe that the furfural formed in the acid solution is in turn

destroyed by the acid, the rate of this reaction being naturally

dependent on the concentration of the furfural. Later results
have also borne cut this theory. It may readily be understood

then, that if one wishes to obtain a maximum yield, he must keep the

concentration of furfural in the acid solution at a minimum. One

way of effecting this is by rapid distillation, the method which

has been commonly used.

 »i<j;
Or «?
•'
1 * . WP^. 'T- ,

. ^ ^

*yTiU ttsifi r.;if?if 0^ (r»vuiA ,j: a)cy:r? j

.*''

iDteAis
^ ’ BKM >1* .n I'i %t^
k
"
.

' '

|i
,.
;'V ri I
J
# »
>;,-•
»f _
‘

.
^ «
* - *
^ *
*
V 1

-0
'

-S 9>t S'-ti f 5?-'i^b|j^ wa j

Cu; I

e;J? ^^;i^*-ia/''K tiguitaigav fftflat: i 4 »t ^<n&'

e^ffiirXiw *C^v
'
‘-
' ••
*
v_'.' '•?!. i

i,-
*^0 tr/jki^.e't it.nevpoti iifii^;|
*5'i q r- ttf.r U; -

Hf-V' "J

rtf 4U' Uii i-fc

•
4^ocio >4 ;Mi'£«o
»
D '
.
a •
-:

^0 hiA^ di( 4 t
8? 0? 1 .|S 6 . 4
‘
l.wf 41
^ J -»vl ^ •

t »'f^J^^a^^^^JI fi»tf tfv ! .P- T I tti . '^X)i iA e.^i^%l.:iC»J» iil

.
.
'

^:
•
- .
:
'^'
r
^ ” '

""m ’
•

^9a- n4,m tfu 4uiuCat A gr^ •*!

'.itot'f-rUy Pi iit7Jo%n.i?t o?

"
ao tA0tvt7-c( » ;©tf? ,
'
iMiism T -rCoiclw -
r .
^

"
4J

' ''
&• f'ii^X}j‘ fe'
\
.

« ^
>.r e« £ -
gjj

fvii*-<r'irLXbXY <sb‘5,x.‘lxe^ Mvo^q

; ;» T
V', '
•

i Kt 3 i,n^Xu -“ i?X /Ji: tws^d^"U X^^-isyijixatf

& F .# *. )•

T\ . -I

.tl/z-t*.? ijq -,-»:::sa Ani'-£S^ "Xdi

;?% "
''•

;# C ,

- -

»>X’.A. • t, :*'lr ;o irtf Wt^^-Xio ^c.?^A9gt4*

I C

^ rt- oflX^ --va.X

41 .-.'f /Wt>XY if. c4tc *i •fj0 \no(li

* '

'

zimiP fi viA figilxtXaa ;ui564 otfX JfX Xaiw^v'SW/^ 'I'o ftC‘f5«ll090aao '

'•^
4i '
^ ^ '

f;od;»^ ml2 '’el ?ialyoen» 'lo 'IAA

T l-o^'irco' A»v.--
 - 13 -

I found th 2.t this sautne result could be obtained by the

addition of a large amount of salt to the acid mixture, for it is

well known that furfural is more volatile from a salt solution

than from pure water. In this way it is driven off with water

vapor at a much lower concentration than when acid is used alone.

This "salting out" process, then, explains why, in the past,

it has been customary to use relatively high concentrations of

acid. Experiment had shown that low acid concentrations gave low

yields of furfural. It may now be seen that this condition was

not due to the lack of hydrogen ion concentration sc much as to

the absence of the "salting out" effect. When salt is used, the

acid concentration may be reduced to l/5 or l/‘6 the previous

value, and it appears that the results are more satisfactory,

the lowerthe concentration. This is just what one would expect,

since, by the use of salt, the acid concentration may be lowered

as wclll as the furfural concentration, and thus still more decrease

the speed of the reaction which destroys the product.

With the knowledge of the above principles and by the use of

proper condensing arrangements, I have been able to distil the

furfural from a corn-cob-acid mixture, separate the oil, and con-

tinuously return the water (together with the furfural in solution)

to the flask where it is used over, thus making a complete cycle

of operations. The process is entirely automiatic, once started,

and the crude furfural (oil) m;ay be drawn off from, the condenser

directly, thus making it necessary to handle no water whatsoever,

and totally eliminating the laborious procedure of concentrating

 \

<r<S
fr -

y i --f

1^ i ? .

>7

IV r'c L
':
*c.
-
rr .

-
i r/f 5
; . .
.'
t» il *- * 1.’^^’ (^-v •

{ i '.. 3
£ . I i.* A« I"-

<£18

V .,-'£ X'
V ^ I T Cl' <1 .

31 ; .•.
’ .•
. T -^ -
i' .' ..
\f '

ot ^.--' - ‘

^ urt^
lie ki i ••-.'.A- >.o-

o- » <5 Cj V -.
'
-*t •
jT

vi- '
jL w - ..^ ^T ' i3VJ4/fe2n V'iii* -
-^ ,-v L>5 V

; Sj-f -
r -' rirr? c.,;..
,
'

.'•e ? ,

?/ ¥*<f T*;: «, ^ 4w 4 '*v

effc-e^OJt; tffr. f .
t.-^D .
i
^ ''
,.“r^-3;3 . L.^l :;T. 1 Sf'l &4

(c ^'C'D'Cit'
•
'
-Vr
-V
.< :•
•
#
, 4

^ C> # -
?
•K *V »“1 •
• - >^,<->

S'

. V-
. r-
•i v Su
.
*''
T (,.
- r.^dUl
1‘Vi/:.* '1
.i X '

• * £-^ " ’tX^^-^‘1 *’‘C V

t' i_ -i '- r *.» L-.\. \ - f ) : ’r'
- I

- h + - l
<--^
;
» - C-* s - o; X .
V* -V - «
:-VJ ai
,.V '“'
*

.Cut- -a'a* >. - - r , ^ _

.
"“;
Ji* n , a- . xC
fv. -
I

-.'*
, it i,
<''’'.
''3
. -V -r
" .<• '
“Cf7 'I
' ;

i.

^-
''*r _
' ' '
i' tV--" r" -
i
^
*, .1

S'i M
 14 -

large vol^amea of iilute solution. Following is a "brsif review

of the experiments which I have male. A number of others which

I had in mind could not be performed for want of time,

ETt^FRIWFNTAL,

In this work all experiments were condiicted with cobs of the

previous year’s crop (1919) of yellow corn. The corn had been

shelled only a few months prior to the date of this research

(Oct. 1, 1930). The cobs were ground in a coal crusher to approx-

imately the size of a pea, there being some pelces considerably

larger and also much fine bran-like material. Sulfuric acid used

was the ordinary C. 95^ acid. Salt used was the crude barrel

product. When weight of cobs is given, it is meant air dry, ground

cobs (about 10*^', moisture), and by percent of H^SO^ is meant acid

of sp .
gr . 1.84,

Expe riment 1 ,- To study the effect of hydrolysis of the cobs,

previouB to distillation by the ordinarv method, 1200 g. of cobs

were placed in a 12-h . flask with 6 L, SSf-* KaSO*, The flask was

placed on a steam cone and heated for 36 hrs., intimate mixture of

acid and cobs being maintained by occasional shaking. The mixture

was distilled with a rapid current of steam, 5 L. of distillate

being collected. On concentration with salt this yielded about

10 g. of oil.

Experiment 3 ,- This exj^eriment was performed in order to

determine if hydrolysis at the boiling temperature would increase

the yield of furfural. 1500 g. of cobs were hydrolyzed by boiling

 •k H ^ y .-I
a: : V -.
,. 10
0 -.'

- ,XP‘ « .

i/,
~r^u . 3 l'
L.

- ^ r.'"

j
.

•' ••» • '

.-.v .I'iOt ' (Lv. ,;

• jr| \-> .
> i r.

V '
» '
1

:
...'•; \
^ ^
'

*
'
'
f
.V ,
•,

'~**J
'
,
y •

V ' •' ’ ' ' r,-- '0 iM • -

... iHL ’-'I i

* 1
,

^ i - j
'
.
*• -<
^
.foO)
^

ij ^6 9/S i'.cetii

••
•
.; _r -.»...,
r-
'- ., 1 z.ig'i,'- .
: r'''.-;ijill'' Cj .

> 4 » i >
" •
'... ,' •
,
" '*1
3'^*.
'>C

.: vO*'' G r.' c< ' '-? wiW 1- ^

:-

-
, Z^::yy V
i

i:% ’
: > '
. . , .V

rf -^0 .^r _'

-’* * ^:^
*“ ^ '
:;-yo
' •
-V. .
j
. 1 V-"' .
.'(1
jj i- r :.

,
SI

' ; i ^
IC 1 .
'

r .3 - J.
, nl '-Vi if&fc'*' 1 . 1
; r! I ' .

- •
: ,
’
i‘«. i i* r-..
*•*' .: -C»iq
:,ig
c. n r
1 ^ •As ..- r i V
'
’r-.
'Bir'

’
'
"
[
• •
-
‘Jt :i ;
'i
’ '
* ,dX"'c.

’* -
.

r •-.:% 1 -
• n- •
•’ . >vi ' .

. : : j . . oi

-.
rri i.i'r.!

•'
. r .*
•
M ^ i 4* V ' » * A

, Li'i ''0..^ : ?
 - 15 -

gently under a reflux condenser v,lth 7 L, HgSO^ for a long

period of hours. The mixture was distilled with steam as before

and the furfural obtained by concentration. Tbe yield was still

less than in the former experiment.

Experiment. 3. - Suspecting the decomposition of the product

by the acid, I made an attempt in this experiment to effect the

hydrolysis with lower acid concentration. 1 Kg. cobs were re~

fluxed gently vTith 5 L. 5^ HgS 04 for a period of 42 hrs. Steam

distillation was then used to bring over the furfural which should

have been formed, and 2-1 /2 L. of distillate were collected. Very

little oil was obtained from this on concentration.

Experiment 4. - In Experiments ( 2) and (3) it was noted that

beads of furfural collected in the reflux condenser during hydroly-

sis. It occurred to me that if this oil could be kept from return-

ing to the acid, the equilibrium between the reaction of formation

and that of decomposition of furfural, might be disturbed in the

direction of its production.

To accomplish this a small trap, shown in Fig. 1, was placed

below the reflux condenser, thru which tdie condensate was required

to flow on its return to the flask,

1 Kg. cobs were placed in the 12-L . flask with 5-L. lOio H 3 SO 4 .

To this was added 2 K'g. salt for the purpose previously discussed.

Heat from a large ring burner was applied to the flask. After 3 hrs
of refluxing, 40 g. crude oil had been collected in the trap and

was removed. 7 hrs. longer distillation yielded yet more product.

mmamtarn
r->*‘

ti . :
^ '

1 .
.. ' w ' ^ ' -
‘
fcVAd n

u
-•'
. ** / -.

f*
A^safiooti 3C :

^ ;'~i .. t-

' ^ X -U' ',‘•1’

' . ^ si."'
*
(

>-rt Tv ., E.1J tail

^ -V ,

. V.
>w:

..r^ '•L :,;70 •'vX- >:?-

.: 'i i 'Z i
'-•
r -

> A,

'

..r-T' ta»H

ti .Xj-'
ItU
'•
»’u'. 1 ¥iT'

X
 :

16 -

These portions were comhlned and distilled and amounted to 93 g.

pure, dry furfural. This is a yield of 9.3'^ of the weight of the

cobs taken.

Experiment 3. - A test showed that furfural is much more

soluble in hot water than in cold, and consequently, a>mong the

ideal conditions to be fulfilled by an apparatus for its removal

from the condensate, is the return of the separated water (satur-

ated with furfural) to the flask at as low a temperature as

feasible. This condition was not fulfilled by the simple trap
(Fig. l) since the returning condensate was kept at the boiling

temperature by the ascending vapor. To correct this, the appar-

atus sho’ifinri in Fig. 2 was designed. It is essentially like the

former but has a short condenser interposed between the trap and the

reflux to.-cool the-^returning condensate and thus precipitate m.cre

furfural. It is also provided with a stopcock below the trap,

,

thus making it unnecessary to take down the apparatus to obtain

the product. With this apparatus, the following experiment was

made

1 Kg. cobs were distilled with 5 L. 10*^ H 5 SO 4 and 2 Kg. salt

for 11 hours.

75 g. of distilled oil were obtained which is a yield of 7.5f^

of the cobs used.

Experiment 6 . - The apparatus used in the foregoing experi-

ment was seen to be inefficient in its cooling of the condensate,

partly because of some steam entering the lower condenser from

above, and partly because of the warm atmosphere surrounding the

trap which must necessarily be above the flask and burner.
To insure
 ^
Tk. WmXL

i"'* <

Q1LS.\^1 . !. J. l‘.i.
'?- '
; •
*J-
- * . e :;x.7

F« 10 .
7 /’ : •:'
[{ Ti^il

'’.1 :t. .

i" ... .
?* k *'

J**
• J. V’’**!'

(i- .’ vT)

;«V 1-.. j:'.,-: yc' - ...

.

'
. -. .i©J '
Bw - "i r, :>H - ! tii4

- ^Ar.i!»trvi> 'ifXdtf'i -M .Ajtd

nu
 .

17

better cooling of the ccnlensate ,

a water jacket surrounding the

trap was seen to be essential. An apparatus including this as

well as other features was designed (as shown in Fig. 3). It

consists of a vapor tube leading to the top of a vertical conden-

ser of large capacity, this resulting from the use of large bore

tubing within the condenser thus offering a large cooling surface.

This form of condenser was preferred to the ordinary narrow tube

type because of the prohibitive length of the latter, cepacities

being equal. The lower end of the condenser, v/ithin the water

jacket, formed the trap, the water of condensation being led out

thru a small tube extending about 5 in. uj: from the end. This

tube was bent thru 160® at the inner end to prevent globules of

'furfural from falling into it and being carried back to the flask.

It was connected by a small rubber tube to a point on the vapor

tube sufficiently high to maintain a liquid level within the

condenser about 2 in. above the water outlet tube. This was done

to allow the condensate sufficient time to become cold before

returning, and also to prevent drawdng off the oil which collected

on the surface. In order to insure atmospheric preseure within

the apparatus, communication vdth the outside was established by

a tube of small bore extending from the top of the condenser down

within the water jacket and connecting to the inner condenser tubs

at a point slightly above the level of the liquid vdthin. A stop-

cock was provided for the removal of furfural from the trap as

before

It is to be noted that this form of condenser is very compact,

having but one water jacket, and the connection to the flask: is
 '
.r'j '?',.
?: ? '.
»<'-#»
l^i’.

/ ^ ti» t

*-
c > ''t

-1

,j.i ,<

n *Nl.
'•..:> 7 or: ^v> » r

.', -* .,-
'
•

„Uy :u-' >5-V '

-
^’•' ‘•' •

*’
Y .^-

••
,
~; >» ••

•> ';:>
i^^0 ».'tsX I (

jtA I

i ,:j
4;

•SIL.:;? .
•• 1 'i
,w -*. A
; i. '.
*
• ’
’s-
-
w
2
''
.« :vt
1
Y, '.- Vfwr 'I i-’S
-^‘iO " :“( ; ( Y ,ii' J -.'"f •- Cl." W ' I

'*
«
,
- A ^ » . «,» »• f

'fe'
’
"‘•.
X. v-r « ' , .
'
J'yk'
.« ' i I
^
'j

. <• V 'tv.
>
- ^
'

. .A ~'l

',> .. .
,-. . cv
' -

t .r
^ r'j *:"v' I.'', i'

'.’•’ .’;•:»
-
-i''
r* ^ 1

'
'

‘"'v

.V .>
: , ,'s; .*o‘ --T ,:.
.

'4' ^A-,y' U‘-

'.n.^Jw f.,!
•{ t

, -
fj l *"v; .

'

-. '. '

„ :’a

'*"
Y

'
i •» - - - ' - ^ .
- »'
)
- ‘
. i *

#r«wcv r' A«<-> .tY i *'

 .

- 18 -

most simple and can be made with the greatest ease and quickness.

In operation it was also found to be very efficient for the pur-

poses for which it was constructed. It was employed in this and

remaining experiments.

1 Kg. cobs were placed in the IS-L, flask with 5 L. 10^ H 2 S 04

and distillation allowed to proceed for 1 hr. No furfural what-

ever separated during this time. 2 Kg. salt was added to the

mixture and ss distillation was resumed, furfural began to collect

at once. Distillation was contained for 7 hrs. luring what time

crude furfural was collected as follows:

First 3 hrs. - 110 g.

Next l-l/2« - 30 g.

" 2-1 /2f’ - 5 g.

This data affords some idea of the course of hydrolysis.

The product was v^ashed with a small amount of lOfb NagCOs

solution followed by water (about 15 c.c. of each) from which it

was separated and distilled under dimini shed pressure. The dis-

solved water with a few c.c. of furfural was distilled off at

first, after which the pure, dry, pale yellow oil was brought

over at 110®

Yield - 124 g. 12.4 ^ of cobs used.

Experiment 7. - This run was a check on Exp. . 6 and was -made

essentially the same.

1 Kg, cobs were distilled for 6-1/4 hrs. with 10^ HgS 04 and

salt as before. Cru.de furfural was collected as follows:

£ -rrX a

^ i i.

'

'
-•;> x'T
 .

“ 19

First 2 - 3/4 hrs. 120

Next hour 20 S*

« 2 - 1/2 hrs. 10 g*

The proluct \7 as neutralized and washed as before and dis-

tilled at 100 ©, almost colorless.

Yield - 133 g. = 13.2^ of cobs.

Experiment 8 . - It was desired to learn if the action of

the salt 'was merely physical (salting out) or chemical, since it

is well knovrai that a boiling solution of HgSO^ and NaCl must con

tain HCl also, and it was of interest to know if the presence of

the KCl was essential or ad‘'/antageous

In thias exp^sriment, the salt was replaced with NasSO^ and

the distillation performed as before. Mo furfural could be

obtained. This was later explained by the neutralization of the

H 2 SO 4 present to form NaHSC^ which, apparently, did not furnish

sufficient acid concentration to effect hydrolysis.

Experiment 9. - To eliminate the above difficulty, this

run was made with the acid sulfate.

1 Kg. cobs were distilled for 8 hrs. with 5 L. 10^ HgS0 4

and 3 Kg. KHSO4 (MaH 804 was not obtainable) . A larger amount

of the acid sulfate was used than when salt was employed, partly

because of its greater solubility making it possible, and partly

tc make the comparison favorable to the acid sulfate.

'

Crude

furfural was collected as follows:

First 3 hrs. - 55 g.

Next 5 ” - 25 g.
 V
V

*
’If f t !»1 ,
.'
tJlW &i '

'r

*
. .
.. •
%

•
’.*1
. J V '- '••* ^

II I M
 - 20 -

Yield of distilled product - 73 g» * 7,3^^ of cobs.

Exceriment 10 . - In order to determine the effect of varying

the acid concentration, this and the next experiment were performed

using 30v^ and H-SO 4 respectively.

1 Kg, cobs were distilled 4 hrs. v7 ith 5 L, 20 f5 H 2 SO 4 and

2 Kg, salt, 75 g, of crude furfural was collected during the first

2-1 /2 hrs. after which no more could be obtained. Much more

rapid carbonization of the cobs occurred than when 10^^ acid was

used, and much foaming ensued which necessitated distilling at a

very reduced speed. In spite of this, furfural was evolved quite

rapidly, more rapi dly, apparently, than in any previous in;in.

Yield - 62 g. = 6.2

Experiment 11, - 1 Kg. cobs were distilled with 5 L, 5^-

K 2 S 04 and 2 Kg, salt for a period of 6 hrs. At the end of this

time, oil was still being evolved but a break in the condenser

caused the discontinuance of the experiment. Crude furfural was

collected as follows;
\
First 2 hrs. - 52 g.

Next 1-1/2” - 43 g,

” 4-1/2” - 55 g.

This was neutralized, washed and distilled and gave a yield

of 125 g. = 12.5^.
 .v-5

.. _ -.vv*. •••' * -

M
• *
»i ' I

'
* VM I. '-i

U V> %
“3 .

j.'C';

• ir'.v
W * , J f'' •
'w -

U«» ;i .“Oi f o fjc ii) i •jx'

..
}
c
.

7 o , :>; ; . . :v CT:' f •> .!-

- V
i-o. i:,-' >!v iyir
' »
.;

t 4 7

^ -<•
a

t’’

J 0 " V,

\'7

- i»‘:ji
i- “wj ,iiC

'
.#
*
r. ;ri^. j;:- i-r
£9i$ i

To
•Ml*
 .

- 21

NOTES

One remarkable feature of the prepa.ration of furfural by

hydrolysis, is the unusual purity of the crude product. Distilled

from a heterogeneous mixture of unknown comp.osition, the oil may

be almost completrely distilled over a range of 1 or 2 degrees,

only a trace of high boiling residues remaining.

Another interesting point in connection 'Alth the purity of

furfural is, that, altho it may be quits colorless when distilled,

it often darkens and becom.es black on standing a few v/seks, or

sometim.es even a few hours. This blackening is probably due to

decomposition of associated impurities with form;ation of colloidal

carbon. The tendency to blacken seems to be largely influenced by

the temperature of distillation, consequently vacuum distillation

is advised. I discovered that furfural dried over CaCls, bladkened

over night and this led m.e to suspect that the presence of acid

on distillation was an important factor. And I found that neutral-

ization of the oil before distille.tion greatly increased its

stability in respect to color. Some so distilled has thus far

as3um»ed only a golden color after two weeks standing, altho it

was distilled at a relatively hish temperature (ilOo )

( 12 )

Beilstein gives a method of purification which consists

of treating the crude oil with small amiounts cf chromic acid to

oxidize impurity. It would be interesting to learn how this

treatment would affect the tendency of the furfural to bla.cken.

 I ,'i

*q?f»

C I
V-V'
t"T.
J
‘A ^ >e‘
.vii ^
'

,*' ’ A '

-. ^^‘

>3(^1 I j^t'
'

^!
f,
t, *;
r^v iff t' ? -Cy :
'
t'lp^ K M i n .sj

k'- pi i-
"'
' '*. I -i
.
‘
V'S'j^ •*'*
^w"" > v
'*' i
’
*
'W^ 4-1^'
^0 vi4p
||

T. ^ ^'r f If-.{# ..®a»t^,

Bei I

Jf .,
"
,mm
'
1 « ^ ^ Ik- ^ilf
',^

,'. '^.,^f^« .tvcj: fr?? 'fc^ 4t^o. ji

> •«
><
i
* ' k

I" >^ih »4l =rfi»l^;^Ccf ?. .« vat i ^»f ^tjaui-ej:

•
-- M a-
'
‘
k.- r% "-^y '"* ,-
''
V - ^ .

^ X4>VxI.^;'^'U .ar/Ss-SKti/A 4l|f

's::^k rui'tv^Z%U \tl‘y'^>: Cti of &stT ^

k ^0
^

^*:* '
H I

TT . •; ~,r'' •
A. ^-*--^ :h

?|» X--:i “ik^. . '.r -C ,OjS ijf*, .

na
^ “"
'

‘
.• •
. ..- .- :,
1-^''

>, •!
” H-- » IT
'
I k
•'
'

•
•*• '

»fii*
*'
'’r'
' '
4
^ ‘.^
^ -i1*t

" * *
j|
i ? i”, • ^
*

^ ^
j

.
.•*
r’^O'Vti. *it .» xV^t aV:^*4r-C*:£)0G fl.opXc3J 'OCfUiJ^O.
'•!J3|a 'ocfuij
I

( . V k J1

6i » ;IM Vi |; 4.^1 mu, Xit> :

"Xvif vis(< :u,i*.*- '>t ^m..s ri»- ‘'^

A*y*3’*' #.^ . ; a?
.
'
V '
^ ,
^'
A'f
:if '»tf 't V* »-» J* ,r
•

. :
->M .
'V -r^m

V 'VI*
,iS
-ja ixneacii

lltl
^ ,.P
V •
 . .

- 22 -

Another Possible ATp-plicatlon of th e Method.

Because of the greater speed of hydrolysis and the probable

increase in furfural yield, it seems quite reasonable to expect

that this new method fcr furfural production 'will have wide

application in the analytical field and appreciably alter the

existing methods for determination of ppentoses, pentosans, oxy-

cellulose, etc. It has been impossible in this research because

of the limited time at disposal, to attempt experiments along

this line

COTJCLUSION

A study of the experimental data makes the follovdng points

seem essential to the preparation of furfural if the maximum

yield is to be obtained.

l) The acid concentration should be as low as possible,,

economy of acid being balanced against economy of time. I have

shown that high acid strengths, altho it may increase the rate

of hydrolysis, tends toward a low yield, and I believe that

somewhere between sf? and IS'^- lies the optimum concentration.

3) The returning furfural solution should be cold in order

to return as little furfural as possible,

3) The returning solution should pass down the tube up

which the hot vapors are rising. This has the tendency to steam

distil the product from the returning -water and concentrate it

in the condenser. The low yield in Exp, 5, indicates this, and I

also h-ave observed in the operation of the condenser in Expts, 6-

 - 23

11 that, aftar drawing off oil from the trap v»rhile the distillatioi|

was in progress, the rate of furfural separation was greatly

diminished, undoubtedly due to the temporary cessation in the

flov/ of returning solution resulting from lowering the level in

the trap.

4) The presence of a salt is quite essential to the oper-

ation of the method as shewn in Extp . 6. KHSO4 may be used in

place of NaCl in the cob-acid mixture, but does not seem to be

as satisfactory. This would indicate that HCl is a more satis-

factory hydrolytic agent than HgSO^. The use of KCl and NaCl

to the total exclusion of would probably give quite satis-

factory results. However, the elimination of KCl might be desir-

able in the technical manufacture of furfural where copper or

other metallic condensers were employed, because of the volatility

of this acid.

5) Because of the volatility of KCl and the formation of

volatile organic acids, (acetic, etc.) during hydrolysi s, the

crude furfural -will always be found acidic. I have shown that

it is desirable to neutralize these acids and leave the product

slightly alkaline on distillation if the color is to be preserved.

SUMMARY

1) A theory has been developed concerning the rats of

formation and decomposition of furfural in sulfuric acid solution.

2) A method has been worked out for the preparation of

furfural which is practically automatic in regard to labor;

 24 -

7/hich is more eoonomical in regard to chemical consumption, re-

quires less time, and is capable of giving a larger yield than

any previously existing method*

3) Several types of apparatus have been constructed and

data obtained from each in operation. From this data, conclusions

have been dravm which may serve as a godde in the operation of

the method, both on a laboratory” scale and commercially.

4) The effects of varying two important conditions of the

process, namely, concentration of acid and kind of salt used, have

been studied .

5) One cause of the discoloration of freshly distilled

furfural has been determined.
 -ri ',<soi^G^v'/&floo -al Xfio^Mor.asa. exoe- s t rfoitfw.., ,

ft^ai ip *©ai^ -e^sl o6i,

fi

. ©t Va its JUV U tiot/ei q

eJCnJ^^qq* to ds-qjft CS

;£rpifti.^iiDil:DC i\^>di^ »p?' .fl,^vti^T«qp a I rifo Afr -coTtt-

.•=r
\
tb Aipl^sifvqo 04ft eY*n!« ifc'JU^r i^TprarlJ :qe^J er4S«
" irX-'Pi if.ii JB ftii
'' **
•'•• •
. ;^r»;
. YjXB-?:o'?e<!UiJOO' i^ao sXjp*! ;£ ^ffo Hto^f ,xpdi^

bXfi C4ft'^Pl"<?iAy?0 fctJJCVl® (^a.

*' .-’ /”'. -
'
'^
- .^.
•?
' ..^ 1
^. ^'i -
^
cvad ^\tii^u tx^c: lo bflrii f/r-Q to- «cXt£ixa»on;op ^ 'iXoiB«q\^«c^piqf
-• '
V -
• "
.
•
^ ^ ,
J
r

I3r
*> .
>T- ^^4 Jf-:
*
/'Vi
^' -^ ''

•
-
” >
‘
w*

V' ;-. YXEfsoat io -^a^tt^’icXobAXft tP ^&x/««or s^O' (^

4 '
_L :
V"-:^ .'.
•"
. •'C
’

" .-SPH
 - 35 “

BIBLIOaHAPHY

l) Organic Chemistry - 1-269

V (C. Nagel - Wochschr. Branw - 30-345
(K. Forster - Ber. - 15-230,323

3) Papierfahr. - 12-1040

4) Dobereiner - Ann. -3-141

5) Fovmes - Ann, 54-52

6) Z. Physiol. Chem . -50-209

7) Cross and Sevan - Papermaking - 21

b) Thorpe - 708

9) Babo - Ann.- 85 - 100

10) Allen - 1-400

11) Zeitsch.- Angsw.Chem. 27-654

12) Handbuch - 3-731

13) Ber. - 34 - 3019

14) Lab, Man. - 190

'
c >*/.
Vi
.-1 \ani
I
3«<33Br
I.

; 1

>. •.