ACID HYDROLYSIS OF CELLULOSE
IN MUNICIPAL REFUSE
A Division of Research and Development
Open-File Report (RC-02-68-11)
U.S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE
Public Health Service
-------
ACID HYDROLYSIS OF CELLULOSE
IN MUNICIPAL REFUSE
A Division of Research and Development
Open-File Report (RC-02-68-11)
written by
Richard A. Chapman, Sanitary Engineer
U.S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE
Public Health Service
Environmental Health Service
Bureau of Solid Waste Management
1970
-------
ABSTRACT
Purpose
Efforts of this research project are being directed toward develop-
ing reaction rates for the formation and decomposition of glucose from
the cellulose in refuse in an acid hydrolysis process at various tem-
peratures and acid concentrations. The rate at which new products such
as 5-methyihydroxyfurfural; levulinic acid and formic acid are formed
is also being investigated. The hydrolysis reaction is being studied at
temperature conditions not previously studied, up to 230°C. The hydro-
lysis process economics will be estimated to determine if it is com-
petitive with other methods producing the same materials,
Methods
A high-pressure, continuously stirred reactor is being used for the
determination of reaction rates. An injection system is being used to in-
troduce reactants once the desired temperature is reached. The reaction
rates for the formation of glucose 5-methylhydroxyfurfural, levulinic
acid and formic acid are being determined at acid concentrations of 0.2,
0 0 0
0.4, and 0.8 percent and at temperatures of 190 C, 210 C, and 230 C.
Results
Preliminary results indicate that in the temperature range of
0
190 C to 230 C, glucose decomposes at the rate predicted by extrapo-
0 0
lation of low temperature (170 C to 190 C reaction rates data. Also,
quantitative saccharifications of refuse indicate that the cellulose
content of municipal refuse delivered to the Center Hill, Cincinnati
incinerator is between 40 and 45 percent.
111
-------
TABLE OF CONTENTS
I. ACKNOWLEDGEMENTS 1
II. INTRODUCTION 2
I II. METHODS 5
High Temperature, High Pressure Equipment S
Analytical Methods for Product Determination 6
Review of Saeman’s Equipment and Analytical Methods 6
Experimental Theoretical Bases and Design 9
IV. RESULTS AND DISCUSSION 14
Initia1 Experimental Design to Elaborate for Simple Solution
of Reaction Rates 14
Revised Experimental Design Facilitates Determination of
Reaction Rates 21
Refuse Potential Glucose Content High Enough For Economic
Consideration 36
V. PRELIMINARY PROCESS ECONOMIC EVALUATION 39
Glucose Price and Market Potential 39
Methyihydroxyfurfural Price and Market Potential 40
Levulinic Acid and Formic Acid Price and Market Potential --- 41
VI. SUMMARY AND CONCLUSIONS 43
VII. FUTURE RESEARCH EFFORTS 44
VIII. REFERENCES 45
IX. APPENDIX I 48
i v
-------
—1—
I. ACKNOWLEDGEMENTS
This project is one of many supported by the Bureau of Solid
Waste Management directed at the utilization of solid wastes. A
study conducted by lonics, Inc. under Contract No. PH—86 -67-2O4
(Conversion of Organic Solid Wastes into Yeast, An Economic Evalu-
ation), provided some useful information related to this project.
I would lik to thank Mr. Robert Fagan of Dartmouth College
working under Research Grant No. EC-00279-02, Kinetics of Porteous
Refuse Hydrolysis Process for his helpful suggestions. Also, the
assistance of Robert Thurnau, William Kaylor, and Raymond Loebker
of the Bureau of Solid Waste Managements is greatly appreciated.
-------
-2-
II. INTRODUCTION
Municipal solid waste contains between 40 and 60 percent paper
which is currently processed and disposed of by incineration, sanitary
iandfilling, or other suitable methods. With the increased use of dis-
posable paper goods. such as limited—wear clothing, hospital apparel,
bedding, and packagingmaterials, the amount of paper in refuse will
certainly increase 1 2 . The development of a suitable and economic pro-
cess for the utilization of waste paper will considerably reduce the
amount of waste that must be disposed of and will reduce the wasteage of
a natural resource. One such process under investigation and being re-
ported here incorporates a weak acid hydrolysis reaction in which cel-
lulose present in municipal refuse is converted to glucose, 5-methylhydro-
xyfurfural, levulinic acid and formic acid.
In the acid hydrolysis process the B-glucosidic linkages of alpha
cellulose are broken and glucose is formed at high temperatupes (above
170°C) and in the presence of a catalyst (sulfuric acid). As the re-
action continues, the glucose is converted to 5-methylhydroxyfurfural
which in turn breaks down to form levulinic acid and formic acid.
Kinetic data for the hydrolysis of wood cellulose was developed
by Saeman 3 for relatively low temperatures (l70°C-l90°C) and for sul-
furic acid concentrations between 0.4 aiid 1.6 percent by weight. He
determined that under these conditions up to about 25 percent by weight
of the wood can be converted to glucose.
-------
-3-
The Madison Wood Sugar Process and various modification of the
basic process have been developed 4 ’ 5 ’ 6 ’ 7 using Saeman’s kinetic data.
By means of extensive recycling (up to 18 cycles) and detention tires of
7 to 8 hours, about 50% of the cellulose in the wood is converted to
fermentable sugars in the modified Madison process. Approximately, 33%
sugar yields have been obtained from the 66% cellulose fraction of wood
fibers. Past efforts have yielded about 50 gallons of alcohol per ton
of wood by fermentation.
Recently, Porteous 8 designed a process for the hydrolysis of cellu-
lose in municipal refuse. His process is designed to operate at 230°C
and with 0.4 percent sulfuric acid. By extrapolating Saeman s kinetic
data, he predicted that about 55 percent of the cellulose in refuse can
be converted to glucose. Refuse contains from 50 to 60 percent paper
whose cellulose content is about 75 percent, therefore, refuse contains
about 40 percent cellulose. If 55 percent of the cellulose in refuse
is converted to glucose, as indicated by Porteous, then about 22 percent
of the refuse weight can be converted to glucose. Since 33 percent of
wood can be converted to fermentable sugar and 22 percent of refuse can
also be converted to fermentable sugar, it can be postulated that one
and one half tons of refuse can produce as much fermentable sugar as
one ton of wood. Furthermore, since one ton of wood produces 50 gallons
of alcohol then one ton of refuse should produce about 33 gallons of
alcohol.
-------
-
In the Porteous plant design, the detention time is about one
minute. This short retention is operable because at high temperatures
the reaction proceeds at a much more rapid rate and only one cycle is
needed as opposed to eighteen cycles in the modified Madison Wood Sugar
Process. Therefore, it appears that the Porteous design will permit
smaller and hopefully less expensive process components for a compara-
ble volume throughput with a comparable production of desired materials.
The Porteous process, appears to be profitable and has the potential
to reduce the amount of solid waste requiring disposal. However, kine-
tic data needs to be developed and verified for the hydrolysis of cellu-
lose present in paper only and in mixed municipal refuse at high tempera-
tures, up to 230°C, before practical process design can begin.
-------
-5-
LII. METHODS
High Temperature, High Pressure Equipment
A two liter Magnedrive Packless Autoclave* is being used for the
hydrolysis kinetic studies. Initial runs determined that a method was
required to introduce the sulfuric acid into the autoclave once the
operating temperature was achieved to estabUsh valid kinetic data
under isothermal conditions. An injection system and a rinse system
were designed and installed whereby the acid and rinse water were nitro-
gen injected. The high pressure reactor was barricaded with sandbags
and the controls rerriotely located to reduce the possibility of injury
to the researcher.
Subsequent reactor runs indicated that the sulfuric acid was re-
acting with the type 316 stainless steel in the reactor and Cr 3 ions
were being formed. The Cr 3 ions were in turn interfering with either
the hydrolysis reaction or with the method for glucose determination,
because very low yields of glucose were found. To overcome this problem
the following action was taken;
O A glass liner was placed inside the autoclave
O A glass sampling tube was installed
• The cooling coil was rer oved
O The therinowell and stirrer shaft were teflon coated
$ A teflon impeller was fabricated and installed.
*Mention of Commercial Products does not imply endorcement by the
U.S. Public Health Service
-------
-6-
Further reactor runs indicated that Cr+ 3 ions were still present
for 10 minutes after the acid was injected and then disappeared. To
eliminate this problem, the acid injection reservoir was lined with
nylon tubing which corrected the Cr 3 ion associated difficulty.
Figures 1 and 2 show the reactor before and after modification,
respectively.
Analytical Methods for Product Determination
A quantitative saccharification procedure whereby all of the cel-
lulose is converted to glucose was used to determine the amount of
cellulose present in paper 9 . Glucose was determtned by the orthoto-
luidine colorimetric method which exclusive for aldohexoses 10 .
5-hydroxymethylfurufral was determined by a colormetric method involv-
ing a reaction with aniline acetate . Organic acids are determined
with a Waters Associates Automatic Organic Acid Analyzer 12 . The tech-
nique utilizes silica gel chromatography to separate mixtures of meta-
bolic acids. Indicator titration permits photometric plotting of con-
centration from a recording photometer. The glucose procedural de-
tails can be found in Appendix I.
Review of Saeman’s Equipment and Analytical Methods
To better understand the comparison of Saema&s results and those
presented here, a brief description of Saeman 1 s methods and equipment
is in order. The experiments on the hydrolysis of wood and the decom-
position of sugars were carried out in sealed glass bombs heated by
direct steam in a rotating digester. Soft glass culture tubes (16 x 150M
-------
7
FIGURE
I
BEFORE
IONS
REACTOR
MODIFICAT
z
I-
0
0
w
0
U,
z
a
1’,
(316 SS)
-------
8
HEATING
JACKET
FIGURE 2
REACTOR AFTER
MODIFICATIONS
CO 4TE
-------
-7-
were filled with the reactants, sealed and placed in the digester.
About 1.5 minutes were required to reach the desired steam pressure/tem-
perature in the digester and an equal time wa required to drop it to
atmospheric pressure.
Sugar analyses were made by the Shaffer and Somogyi method 13 using
their developed reagent and a 30 minute boiling time. linhydrolyzed car-
bohydrate material was determined by subjecting the residue to a quan-
titative saccharification followed by sugar analysis.
Experimental Theoretical Basis and Design
Saeman showed that the hydrolysis of cellulose is described by the
following consecutive first order reactions.
k 1 k2
A ‘B
where: A = Cellulose
B = Glucose
C = Glucose decomposition products
and k 2 = Reaction Rates
The amount of A, B, or C present at any time during the reaction is
described by the following formulas 14 :
A = Aoe_k lt Eq. 1
— ° (e 1 —e 2 ) Eq. 2
k 2 —k 1
C = Ao [ l + 1 (k 2 e 1t k e 2t )] Eq. 3
k 1 —k 2 1
-------
- 10 -
where: A 0 = the initial concentration of cellulose
t = time in minutes after the reaction begins
The time at which net glucose production reaches a maximum is
given by:
t = in k 2 -ln k 1
k 2 —k 1 Eq. 4
The maximum net glucose yield is given by the formula:
Bmax = Ao 1(2 k 1 - k 2 Eq. 5
Temperatures of 190°C, 210°C, and 230°C with acid concentrations
of 0.2, 0.4, and 0.8 percent by weight were chosen as the conditions
at which the hydrolysis reaction would be studied. The lower tempera-
ture and acid conditions were chosen because previous research on the
hydrolysis of wood cellulose 15 indicated that hydrolysis temperatures
less than 190°C with acid concentrations less than 0.2 percent did not
produce significant glucose yields. The upper temperature and acid
concentrations were chosen because extrapolation of low temperature
reaction rates indicated that with hydrolysis conditions more severe
than 230°C and 0.8 percent acid, the time to maximum net glucose yield
was less than one minute which would not allow for adequate process con-
trol.
Previous research efforts 6 ’ 17 have shown that sulfuric acid,
-------
—11 —
hydrochloric acid or phosphoric acid can be used as a catalyst in the
hydrolyses process. Sulfuric acid was chosen as the catalyst because
it stimulates greater net glucose yields than phosphoric acid and
although comparable net glucose yields can be obtained using either
sulfuric acid or hydrochloric acid, sulfuric acid is less expensive arid
less corrosive.
The initial experimental design involved the solution of equation
2:
E 1 1 k (e_klt -e -k 2 t)
Kraft paper and Whatnian No. 2 filter paper that had been ground in
a Wiley Mill to a particle size less than 0.5 nm were chosen as the cel-
lulosjc materials to be initially hydrolyzed. A quantitative sacchari-
fication was performed on the samples to determine the amount of cellu-
lose present. The samples were then hydrolyzed at different temperatures
and acid concentrations. At various times (t) during the reaction, the
amount of glucose (B) was determined. The initial concentra ion of cel-
lulose (A 0 ) was knoi.Jn from the quantitative saccharification. The re-
action rates k 1 and k 2 can then be found by fitting the least squares
curve for B through the two points. As Atkinson and Hunter 18 proved,
this is a very efficient method of determining k 1 and k 2 . Fbwever,
sampling times and glucose content must be determined very accurately
for optimum determination of K 1 and k 2 . Duiing the initial trial runs
some difficulties were experienced with the glucose determination due
-------
— 12 —
to the presence of Cr 3 ions in the hydrolysate. The Cr 3 ions in the
hydrolysate came from the type 316 SS acid injection system which was
subsequently lined with nylon tubing to overcome the problem.
After correcting the Cr 3 ion difficulty, the experimental design
was changed to permit close monitoring and better understanding of the
reaction and more confidence in the results. The first protion of the
revised experimental design involved the study of the rate at which
glucose is converted to 5-methyihydroxyfurfural, k 2 in Equation 2.
The hydrolysis of glucose is described by the formula:
B = B 0 e t Eq. 6
where: B = the amount of glucose present at time t
the initial amount of glucose
t = time, minutes
k = reaction rate.
To determine the value of k at each terperature and acid concen-
tration, glucose was injected into the reactor and samples were taken
at various times and analyzed for glucose content. A plot of the
logarithm of the remaining glucose versus time is a straight line
whose slope is the reaction rate k 2 .
The second portion of the revised experimental design will involve
the study of the hydrolysis of the filter paper. At each temperature
and acid concentration, he filtLr paper is to be hydrolyzed and
and samples taken at various times and analyzed for glucose content.
Equation 2 can then be solved for k 1 since A 0 , initial cellulose
-------
- 13 -
content, is known from the quantitative saccharification procedure and
k 2 is known from the hydrolysis of glucose. The amount of 5-methyihy-
droxyfurfural, levulinic acid and formic acid produced by the hydrolysis
reactions will be monitored to aid in the determination of the rate at
which they are formed.
The next phase of the research work will determine the effect of
paper concentration and particle size on the production of glucose and
other compounds. Once this has been found, samples of refuse will be
hydrolyzed to determine if the reaction rates vary from those associated
with filter paper hydrolysis.
To determine the amount of potential glucose contained in refuse,
samples are periodically taken whenever the Center Hill Laboratory
haninermill is operated. These samples are then ground in the Wiley
mill and exposed to a quantitative saccharification procedure whereby all
the cellulose is converted to glucose. From this information, a good
estimate is obtained of hydrolyzable cellulose in municipal refuse.
-------
- 14 -
RESULTS AND DISCUSSION
Initial Experimental Design to Elaborate for Simple Solution of
Reaction Rates
Table 1 and figures 3 through 7 show the glucose yields versus time
obtained by the hydrolysis of filter and Kraft paper at various tempera-
tures and in the presence of 0.2 percent by weight sulfuric acid. Also
shown are the glucose yields from the extrapolation of Saeman’s reaction
rates
In each case, the experimentally determined glucose yield was less
than that predicted by the Saeman reaction rate extrapolation. Also,
in each experimental, determination, the time to maximum glucose yield
was less than expected from the extrapolation of Saeman’s data. There
+3.
were two reasons for these descrepancies. One was that Cr ions were
injected into the reactor with the sulfuric acid and interfered with
the glucose determination. This resulted in a deceivingly low indica-
tion of glucose present. The second reason for the descrepancies was
that the cellulosic material was placed in the reactor and heated for
as long as three hours before the operating temperature was obtained.
During this time the cellulosic material was being thermally degraded.
Therefore, less cellulose was a ailable when the acid was injected and
the hydrolysis reaction began, The heating also made the remaining
cellulose more susceptible to the hydrolyses reaction and resulted in
higher than expected initial yields of glucose and a shorter time to
maximum yield.
-------
- 15 -
TABLE I - GLUCOSE YIEL1 FROM KRAFI PAPER AND WHATMAN No, 2 FILTER PAPER
DATE PAPER PAPER TEMP. ACID TIME GLUCOSE SAEMAN’S
TYPE CONC. (°C) CONC. ( j ,) YIELD GLUCOSE YIELD
(g/l) (%by wt) (%) (%)
3/12/69 Kraft 1 190 0.2 10 21.4 9.5
56 27.4 32.0
90 19.0 33.8
107 14.9 32.5
120 11,6 31.0
3/13/69 Kraft 1 210 0.2 2 17.4 14,0
7 31.2 34.0
9 32.2 39.0
12 32.0 41,5
15 28.4 42.4
20 20.4 40.4
3/18/69 Whatman 10 190 0.2 10 17.5 9.2
No. 2 56 31.5 32.2
i1ter
paper 90 27.9 34.0
107 25.5 32.8
120 23.4 31.1
3/19/69 Whatman 10 210 0 2 2 15.5 14.5
N 0 . 2 7 24.6 33,7
Filter
Paper 9 26.3 38.0
12 27.8 41.2
15 27.8 42.0
20 25.0 40.0
3/27/69 Whatman 10 230 0.2 2 19.1 47.0
No, 2 3 18.5 50.5
Filter
Paper 4 16.1 48.7
6 10.5 39.1
10 3.6 18.8
15 1.6 6.6
20 1.0 2.2
-------
16
FIG YF?
HYDROLYSIS
19o°c I /J piqp
)
OF K S9F7
2Z /L 7 Uz Oq
60
3
30
(3s MIht, t 7.)
0
/
2 ’
/
/
LU
0)
0
c i
I0
/
/
/
Lff6 o
0
0
E (PERI W NTftL.
2o
E P TED yiELD
yIELD
80
I00
TIME (wtIjiur s’
-------
17
F1cuR /
I4yDr o&y , oF !RAFT PAPE
2/bc) / /- PAPERJ 0.2 ,t ‘JT, I/i SQy
Ito
/
,
/ (som&,, 33Y )
/
30
4’
U i
/
/
•1
/
/
END
c 4 j / P eIMENTAL yl&LD
— —— E PECT6O ‘/PEL.O
0
10
TIME C,n, oris)
-------
18
FIGURE 5
HYDROLYSIS of f1LbTEf PR PER.
I 1O•C) Ic ’ ;j Q PR PER, o.2Z
( T/ , 1 IlL S°jf
i51
3’,
( z irnW., 32.S )
Q
>‘
/
‘I . ,
‘I ,
0
U
4
cj
/
/
10
/
/
/
kE END
a
- xPE# jmE 7AL
0
EAPECT D
Y,EL D
YIELD
‘fO
TIME (ma - s)
0
Ba
/00
-------
19
6
HYDROL’/S!S OF FiLTER PAPER
Z/o°C , ic,/ Q P Pff 1 c.zZ A*ir H So, 1
— -
,
/
/
/
10 L.EGE o
E ERlmE,.-rPL .y,E,_O
/ — — PECTEb yiEs O
b
0 10
TI M.E (mi.’ur )
-------
20
FIc JRE 7
U’OROL.,ys,s oF F!i-TE L PftPER
23oC 1o/4’ PqP J o.2 r /h SO,
5o
II’ \
\
I \
I ‘
I
3o1
0
I \
if \ _
EARIThE&Tfl&. yi6I
0
TImE (miwrE )
-------
- 21 -
Revised Experimental Design Facilitates Determination of Reaction Rates
To provide a more complete understanding of the hydrolysis reac-
tion and more confidence in the results, the experimental design was
changed to study the hydrolysis of cellulose and glucose separately.
During the first portion of the study, the rate at which glucose hydro-
lyzed under various conditions of temperature and acid concentration
was studied. The second phase of the revised experimental design, the
hydrolysis of cellulose, is just beginning and the results will be pre-
sented in a future report.
Figures 8 through 16 show the various conditions and rates of
glucose decomposition. A summary of the experimentally determined rate
constants is given in table 2. Also shown are rate constants calculated
from equation 7, the formula derived by Saeman to describe the hydroly-
sis of fermentable sugars at temperatures in the 170)C to 190°C range.
- 14 1.02 - 32,870
k—2.39x10 C e T
Where Cs = concentration of sulfuric acid, %
R = gas constant, 1.9865
T - absolute temperature, °K
Figure 17 is a plot of the logarithm of the reaction rates calcu-
lated from equation 7 as a function of the reciprocal of the absolute
temperature for each of the three acid concentrations. Also shown are
the experimentally determined reaction rates in the ‘190°C to 230°C range.
-------
— S
16,. stt v
J)L’- ‘ ‘1F sjj .,j c r lyre ‘. A . ?. :
‘ &1 (a ? S
( I ’
1 %
p . .
‘ ‘S
‘ S
4
c 4
4 ,
4
I
IS
I S
4 ’
‘I
:4
S
-4
5
- L ‘er: j_ils,_If . ’t • 6P.1rU
1 J : s ilt
r• • , .9... r
a e — r . “4
I--.-- “ 4
I_a : —: L •
f _ s • 5%
pa • V - 4
a
-------
p 4)
a
S
I
• •‘. ,. 1 c adek
a t
—V — ..
• . 1
*
• 1f t
IC
s . C. . ••
‘ p
1% a •. U
— A • at
‘ I
1 %
‘a
S
S
I
‘ I
1 %
•tp
“ U I
‘p
‘
I’
t% ‘at’ •(
. 1 1.
IS
C
a
i i .
‘4 4 •
a .
t 4.
p
‘ S
N
‘ a..
1• • • s
r .7
a ,
* • — It.
U ’
‘ L! ‘ ‘I i
I I
‘$2) ....a. 0 £ •$ i
a ’
.-t #ifljU
‘a .
-------
th e. ? p 1 /.‘
101 e
I C.
‘a
S
. 5
1 %
‘S
*
a
‘I I.
3
• .
V
I I
I t
a
•t
p
• i# . p
‘I
S .. • I — I
I
S
¼ — • — ‘ - t. 7 S
.J , ._:)I e%j r. ,
4 —c btpf! •tf ‘ ‘
4
S.. .) ‘ce;.-.( /;h. kI
,,t 4. p.
4
a.. 1 4.
-------
t a
3’, ’ i i
r” eTjta.
•‘ • Ot ’ a
f t 1 • c ‘ lW ,I.4
N
‘ l b
N
N
b
‘N
S
‘a
¼
“I .
a
-. I ‘r- A f: ; A
V . p
p / l i ; ; r
a I a • P
1 CV r ? . ’ ’ £ •t1
.d ‘f in
. 1. ; a 7• ) e. g •
.4Ii / 1 “otT
0 . - . a— - .. . . ..
1 S
I ,‘Y I . ‘:.‘;)
-------
‘
“sb m a O ..-r q, •.‘
C .
‘4
“ I
I ’
a l.
“ p
‘a ’
C s ’
.4 ) .4
‘4
S.
I ’
‘S
5 Ii. . S ‘ • •
•IP f ! r 1 .L.
V ‘ S
— .
: -• v S j • ft
‘ . f: - : .- I A.
‘ k . “b s , . . )
L t ;,,’
tt ‘ c;..c_
a.
YEA
-------
a •
j •‘t .. Ci. a ) 4 r
-r cc .nP ,7;g b
a la ’ •a - .. F .
• I . ,
4 ’ • 3 a.
‘C
C
N
‘S
‘S
S
‘ I ,
I
‘I
.4¼L 4 ,
L .’
a a • ‘ a
•1 f”; , )4)j eAd % I
a ), alae 4 & t i .OS’ . .
‘4 ’ :. ns 1 h 5’e ’ , . a,
I • - - - - . V. . --..
t ’ ’ C ’ ,J
-------
(‘ jc_ red.
V t? P ( , , ,hMe .i,P
e 9 ’ia C . , ‘ZsCn .
c
t
0
‘N
9;
‘S
( .. .
/ 1 . • £ .• hi ’
1’.; . m e t t g
A_I oaat “ , “ ‘‘‘!
14 ’ • ,;a.’.
A t’ut$ ’U .
4 / .
I ‘ % ‘.t C M I
-------
‘ - p
1 . 1j 1 •P& 1 .1 - f
s.rt
a &1a ,r’’J c • • .. .
.i . .. .., :
‘.dt
lg
4
N
‘S.
‘ .11-
1 ’
S
b
S •
t y r — S
I d S
a .1
£ C ’
SJr :
/ ,. ,I / : - , ,
‘ ‘ I i.-JJ
- E .” ,jz ,.r)_.s&(. ,e,.
4 4 .e stc c a)
S
•
I -p I 4 . 1 - I •‘ a
9_ — I
• ,t% , .. .%. .3)
-------
4
‘ I ,
lb
S
I
4
I
.4. 1 • e j ; U ,
a
‘a
‘ I ’
1’ — C ’ 4 % ‘ I ‘ C d
d .C
bofliln :jIe
I I • i .
4
4
,4.
‘ I
I
1
S
S .
I
S
• 1
it. t.• ,4
a .
,16cs p - ( • .
I p
a
•
‘:
C
e :a $
t
I ’
f
4
I t
:11 ’
:
—
L
41 S
ti
,‘
.
I
•
C
øb
• 1 V •
I ,
‘S
S
I
a
S
\ 4
S
p
It
• • _ .4 (
-------
- 31 -
TABLED - DECOMPOSITION OF GLUcOSE iN 0.2%, 0.4% nd 0.8% SUL-
FURIC ACID AT VARIOUS TEMPERATURES
H 2 S0 4 Temperature (minE 1 ) (min T)
0.2 190 0.0146 0.0142
210 0.0682 0.0621
230 0.256 0.242
0.4 190 0.0309 0.0287
206 0.0915 0.0946
230 0.470 0.491
0.8 195 0 .0723 0.0852
212 0.287 0.294
226 0.732 0.765
-------
32 /
0 - i8o
F1 uRE 17
oI DF RE4 .flc
\ c oMsrF Nr A IN
\ D Composj r,oN of- 6’..vcose
\ \ /J s , - r VA Rio S 5 7 E ’J 6 r i -i S
Pci A TED fR 4 pt Oft TR
Rr /?o°c — /9o
\
L \ \ \ EicPERa,ne NrALL)/ OemI#IEO
\ \ \ cnow coN rAHrs gr
\ \ \
Q QIC
0
0 -
o -
12 -p
‘J O.0Sô
t __p e 0
•
o O.
L
o.oI
ç, miq*is
4
R cJPRc2C4, of IWE F ?6rOJ-(.)rE rnpE, ’j TL1A?t
( K0 A)O )
-------
- 33 -
It appears that there is no significant difference between the
experimentally determined reaction rates and the reaction rates calculated
from equation 7.
A rigorous statistical comparison of the results will be conducted
to determine the necessity of further experimental efforts related to
the decomposition of glucose.
As indicated earlier, Prteous 8 recently designed a hydrolysis pro-
cess for the production of glucose from cellulose in municipal refuse.
The operating conditions of his process were 230°C with 0.4 percent
sulfuric acid concentration. The kinetic data used by P 0 rteous was
extrapolated from the work done by Saeman in the 170°C to 190°C range.
It appears that the extrapolation of Saeman’s reaction coefficients for
the decomposition of glucose gives valid values at higher temperatures.
Therefore, it seems reasonable to accept the reaction coefficients used
by Porteous for glucose decomposition.
The reaction rates used by P 0 rteous for the formation of glucose
are also extrapolated from the work Saeman did on the hydrolysis of wood
chips. It is quite possible that refuse associated cellulose hydrolyzes
at a faster rate than is predicted by the extrapolation to 230°C of the
wood hydrolysis work at 170-190°C. Therefore, reaction coefficients at
temperatures up to 230°C will have to be determined experimentally for
final process design.
The production of 5-methylhydroxyfurfural was monitored as the
glucose hydrolyzed at various conditions. Figures 18 and 19 give an
indication of the effect of temperature and acid concentration on the
-------
34
EEF’E r r F ftTz E
s
-4
o 4 z
- -
I
/2o
-------
35
oi
:
1 - - .
ftc u L2N 7 ft rn, J
£- / 7 fl?YL 1iYO øA) fQi FCJ AL
—-
-—4-- —----
—4
-------
- 36 -
5 methy1hydroxyfurfura1 y téld. The reaction rates governing it produc-
tion and decomposition have not yet been fully determined. However, in
figures 18 and 19, it appears that the 5.methylhydroxyfurfural is present
in the hydrolysate for an adequate time to facilitate its removal. As
indicated in figures 18 and 19, the greatest 5 methy1hydroxyfurfura1
ytelds were obtained at 230 0 C with a 0.2 percent sulfuric acid concen-
tration. Therefore, it would appear that operating conditions could
exist for the near maximum production of both glucose and 5-methyihydroxy-’
furfural at a relatively high temperature with low acid concentrations.
Also it should be possible to choose the proper conditions for the pro-
dUction of maximum amounts of both 5-methyihydroxyfurfural and its decom-
position products, namely, levulinic acid and formic acid.
Refuse Potential Glucose Content High Enough For Economic Consideration
Table III shows the results of the quantitative saccharification of
ground refuse whereby all the cellulose was converted to. glucose. An
average yi’eld of 40 to 45 percent was obtained. These yields are
reasonable in that refuse contains 50 to 60 percent paper and that the
cellulose content of paper is between 60 and 80 percent.
Porteous calculated that the hydrolysis of refuse containing 45
percent cellulose and the subsequent fermentation of the glucose would
yield a net profit of $4.21 per ton of input refuse. The Porteous
estimated net profit allows $4.50/ton for the proper disposal of all
nonhydrolyzed refuse by sanitary landfill. Fermentation plant liquid
-------
- 37 -
TABLE ill — POTENTIAL GLUCOSE CONTENT OF REFUSE
Approxi-
Number of mate Mean
Sample Date Determi- Sample Glucose Standard
Number Ground nations Size Yield Deviation
(9.) ( 5)
19 D 10/10/68 5 0.5 38.1 2.36
26 E 10/19/68 5 0.5 45.5 3.24
11 D 10/30/68 5 0.5 43.7 1.95
11 D 10/30/68 5 3.8 39.3 1.61
( I
-------
- 38 -
effluent treatment has been considered by Porteous with allowance of
3ft/pound for BOD load reduction. Based on our preliminary work and
Porteous financial estimates, ft seems possible that the hydrolysis
process would be economical in locales where the cellulose content of
refuse is similar to that delivered to the Cincinnati Center Hill
incinerator. The process economics could be more favorable if the
hydrolysis process is designed for the production of 5 methylhydroxy-
furfural or levulinic acid and formic acid as well as glucose.
-------
- 39 -
V. PRELIMINARY PROCESS ECONOMIC EVALUATION
Glucose Price and Market Potential
Porteous 8 showed that glucose produced by the hydrolysis of
refuse is competitive with other raw materials in the production of
ethanol by fermentation. Unfortunately, only about 17 million
gallons (about 5 percent of total) of ethanol are produced annually
by fermentation’ 9 . Using the Porteous design this is equivalent to
the ethanol production from about 5 hydrolysis - fermentation plants
processing 250 tons/day of refuse. Therefore, it appears that other
uses for the glucose produced from the hydrolysis of the cellulosic
portion of refuse must be found.
Recently Fogan 20 determined that the acid hydrolyses of refuse
will produce glucose at 2-3 cents per pound which is competitive with
blackstrap molasses provided the plant processes at least 350 tons/day
of refuse containing 50 percent paper. Understandably, larger plants
processing refuse with higher paper contents could produce glucose at
an even more competitive price.
In 1967 approximately 300 million gallons of blackstrap molasses
were used in the industrial production of drugs, citric acid, vinegar
and ethanol 20 . This indicates that about 1 million tons of such sugars
were consumed in processes that could equally as well use the product
of a hydrolyses plant 21 . This is equivalent to the glucose output of
about 30 hydrolysis plants processing 500 tons/day each of refuse with
a 50 percent paper content which is equal to approximately 3 percent of
-------
40 —
the United States daily production of 500,000 tons 1 of municipal
refuse.
Methyihydroxyfurfural Price and Market Potential
Methylhydroxyfurfural is formed by the hydrolysis of glucose and
is present in the hydrolysate at the time of maximum g1ucôsie.’ yield.
Further hydrolysis results in an increase in the methyihydroxyfurfural
yield.
Methyihydroxyfurfural is polyfunctional and can be used to syn-
thesize a wide variety of compounds. It behaves like a ivormal primary
alcohol and, in some instances, as an aromatic aldehyde. Ring reaction
include addition, ring clevage and ring clevage followed by closure to
give 6—membered hetrocyclic rings, Little attention has been given to
exploiting this interesting compound, probably because of the difficu1ty
of obtaining sizable quantities of high-quality material 21 .
As stated earlier, about 22 percent by weight of refuse can be
converted to glucose, Figures 18 and 19 show that at 230°C and 0,2
percent acid concentration, that about 120 mg of methylhydroxyfurfural
is formed from 500 mg of glucose. This is equivalent to about a 24
percent conversion. Therefore, if 22 percent by weight of refuse can
be converted to glucose, and 24 percent of the resulting glucose can be
converted to methihydroxyfurfural, then it follows that about 5 percent
of the refuse weight can be converted ‘to methyihydroxyfurfural,
But, little attention has been given to exploiting methylhydroxy-
furfural because of the difficulty in producing sizeable quantities of
high quality material. If sizeable quantities can be produced by the
hydrolysis of the cellulosic protion of refuse then it is possible
-------
- 41 -
that many uses for it might be found. The current consumer price for
methyihydroxyfurfural is $14.75 for 5 grams hich is equivalent to $83
per ounce. Undoubtedly, large quantities cost less per gram and if
sizeable quantities become available the price will drop. However, it
appears that the production of methylhydroxyfurfural from refuse would
be profitable.
Levulinic Acid and Formic Acid Price and Market Potential
Levulinic acid and formic acid are produced by the hydrolysis of
glucose with methyihydroxyfurfural as an intermediate product. About
0 percent yields are obtained from glucose using sulfuric acid as the
catalyst. Therefore, if 22 percent by weight of refuse can be conver-
ted to glucose and 40 percent of glucose converted to levulinic acid and
formic acid then about 9 percent of refuse by weight can be converted
to these acids. This is only a rough approximation and is true only
if the glucose is recovered at the time of maximum yield and then hydro-
lyzed to levulinic acid and formic acid. The actual yield would be
higher than the 9 percent indicated because there is levulinic and
formic acid present in the hydrolysate at the time of maximum glucose
yield and would therefore add to the 9 percent yield previously men-
tioned.
The 1967 production of formic acid was 18 million pounds and sold
at about 15 cents per pound 23 . Formic acid has many industrial appli-
cations including its use in electroplating, dying and processing of
textiles, and in the production of formates and esters including cellu-
lose asters 24 .
-------
- 42 -
Despite the low cost of the raw material (glucose) that can be
used to produce levulinic acid, it has not been marketed as an inex-
pensive chemical intermediate. The controlling economic factor
appears to be the high cost of recovery from dilute impure solutions 21 .
Levulinic acid is produced in small quantities and is available from
major chemical suppliers at about $3.50 per pound 22 .
Once the problem of product recovery is solved, and levulinic
acid is available at a lower cost, many uses for it should materialize,
for levulinic acid reacts to form many interesting heterocyclic com-
pounds 21 . Projected uses for levulinic acid that appear to have
merit include the production of sebacic acid and nylon type polymers 21 .
There is one use to which levulinic acid can immediately be put. It
has been found that its sodium salt has ideal properties as an anti-
freeze agent. It has definite advantages over ethylene glycol for
this purpose. It is a water soluble solid and is therefore more
easily marketed than the liquid glycol. It is less corrosive to the
iron parts of internal combustion engines than is tap water itself
and has no detrimental effect on the rubber connections used in
engines 25 .
-------
43
Vi. SUMMARY AND CONCLUSIONS
The preliminary results are as follows:
1. At temperatures between 190 0 C and 230°C and in the presence of 0.2
to 0.8 percent sulfuric acid, glucose decomposes at the rate predicted
by the extrapolation of reaction rates determined at temperatures be-
tween 170°C and 190°C and in the presence of similar acid concentrations.
2. Refuse delivered to the Cincinnati Center Hill Incinerator contains
about 40 to 50 percent cellulose that can be converted to glucose.
3. The process designed by Porteous for the production of glucose from
the cellulosic portion of refuse and its subsequent fermentation to
ethanol is realistic.
4. Glucose can be produced at a price competitive with blackstrap
molasses by hydrolyzing refuse associated cellulose.
5. About 3% of the nations 180 million tons per year of municipal
refuse can be hydrolyzed to produce a quantity of glucose equivalent
to the annual consumption of blackstrap molasses.
6. Methylbydroxyfurfural, levulinic acid and formic acid formed by the
hydrolysis of cellulose in municipal refuse are of economic value and
if sufficient quantities are produced the price will probably become
low enough to promote their extensive use in the chemical industry.
-------
- 44 -
VII. FUTURE RESEARCH EFFORTS
The following tasks are planned for the completion of this research effort:
1. The determination of reaction rates governing the formation
of glucose from cellulosic materials by the acid hydrolysis process at
temperatures between 190°C and 230 0 C and in the presence of 0.2 to 0.8
percent sulfuric acid. This task should be completed by the end of the
third quarter of Fiscal Year 1970.
2. The confirmation of preliminary results indicating that be-
tween 190 0 C and 230°C Glucose decomposes at a rate predicted by the
extrapolation of reaction rates in the 170°C to 190°C temperature range.
This task should also be completed by the end of the third quarter of
Fiscal Year 1970.
3. The determination of reaction rates for the formation of glu-
cose by the hydrolysis of the cellulosic portion of municipal refuse.
This phase of the project should be completed by the end of Fiscal Year
1970.
4. The determination of reaction rates governing the formation
of hydroxymethylfurfural, levullinic acid and formic acid from the
cellulosic portion of refuse by the acid hydrolysis process. The pro-
jected completion date for this task is by the end of Fiscal year 1970.
5. The study of different pretreatment methods of municipal refuse
and its effect on the production of glucose, hydroxymethylfurfural,
levulinic acid and formic acid, The completion date for this task is
predicted to be by mid Fiscal Year 1971.
-------
- 45 -
REFERENCES
1. Richard D. Vaughan, Management of Solid Wastes from Hospital -
Problems and Technology”, Use and Disposal of Single Use
Items in Health Care Facilities, Report of a National Con-
ference, Dec. 4, 1968, National Sanitation Foundation, Ann
Arbor, Michigan.
2. “The Role of Packaging in Solid Waste Management 1966 to 1976,”
Bureau of Solid Waste Management Publication SW-5C, DHEW,PHS,
CPEHS, ECA, BSEM, Rockville, Maryland, 1969.
3. Saeman, Jerome, “Kinetics of Wood Saccharification,” industrial and
Engineering Chemistry , 37, 43 (1945).
4. Harris, Elwin E. and Beglinger, Edward, “Madison Wood Sugar Process,”
Industrial and E ineering Cherilistry , 38, 890 (1946).
5. Gilbert, Nathan, Hobbs, I. A. and Levine, J. D., “Hydrolysis of Wood
Using Dilute Sulfuric Acid,” Industrial and Engineering Chemistry
441 1712, (1952)
6. Harris, Edwin, et. al., “Hydrolysis of Wood,” Industrial & Engineer-
ing Chemistry 37, 12 (1945).
7. Plow, R. H., et.al., “The Rotary Digester in Wood Saccharification,”
Industrial & Engineering Chemistry , 37, 36 (1945).
8. Porteous, Andrew, “Toward a Profitable Means of Municipal Refuse
Disposal,” ASME Publication 67-WA/PID-2 (1967).
9. Jerome F. Saeman, Janet L. Bubl and Elwin E. Harris, “Quantitative
Saccharification of Wood and Cellulose,” Industrial & Engineer-
ing Chemistry , 17, 35 (1945).
-------
- 46 -
10. Glucose Determination by the 0-toluidine Method, Described in
Nature , 183, 108 (1959) and modified by Dr. Sanka, Mary Hitch-
cod Hospital, Hanover, N.H. for work in blood and urine sugar
analysis.
11. 0. I. Miller and David Liedirman, “Determination of furfural in
Petroleum Stocks,’ Analytical Chemistry 27, 42 (Nov. 1955).
12. Automatic Organic Acid Analyzer, Waters Associates Inc., 61 Foun-
tam Street, Framingham, Mass.
13. Shaffer, P. A. and Somogyi, ,i., “Copper - lodometric Reagents for
Sugar Determination,H Journal of Biological Chemistry , 100, 695
(1923).
14. Frost, Arthur A. and Pearson, Ralph G., “Kinetics and Mechanism,”
John Wiley and Sons, Inc., New York, 1953.
15. Alfred J. Stamm, “Wood and Cellulose Science,” The Ronald Press
Company, New York (1964).
16. Elwin E. Harris and Albert A. Kline, “Hydrolysis of Wood Cellulose
with Hydrochloric Acid and Sulfur Dioxide and the Decomposition
of its Hydrolytic Products.” Journal of Physical & Colloid
Chemistry, 51, 1430 (1957).
17. Elwin E. Harris and Bell G. Lang, “Hydrolysis of Wood Cellulose and
Decomposition of Sugar in Dilute Phosphoric Acid,” Journal of
Physical & Colloid Chemistry , 51, 1430 (1957).
-------
- 47 -
18. Atkinson, Anthony C. and Hunter, William G., “The Design of
Experiments for Parameter Estimation,” Technometrics , 10, 271
(1968).
19. Rickles, Robert N., “Future Chemical Growth Patterns,” Noyes De-
velopment Corporation, Pearl Rwer, N.Y.
20. F’ gan, Robert D., “The Acid Hydrolysis of Refuse,” Unpublished MS
thesis, Dartmouth College, Thayer School of Engineering,
Hanover, N.H.
21. Browning, B. L., “The Chemistry of Wood,” John Wiley and Sons,
New York (1963).
22. Aldrich Chemical Catalog 14, 1969-1970, Aldrich Chemical Co., Inc.,
Cedar Knolls, N.J.
23. Commodity Year Book 1968, Commodity Research Bureau, Inc.,,
140 Broadway, New York, NY
24. Snell, F. D. and Sriell, C. T. Directory of Commercial Chemicals,
third ed., D. van Nostrand Co., Inc., Princeton, N.J., 1962.
25. Wiggins, L. F., “Utilization of Sucrose,” Advances hi Carbohydrate
Chemistry , 4, 239 (1950).
-------
- 48 -
APPENDIX I
-------
- 49 -
GLUCOSE, o-toluidirie
1. Pipette, with a rubber bulb, 6.0 ml. of o-toluidine into a test
tube.
2. Add 1 ml. sample, mix well.
3. Place in boiling water bath for 10 minutes, cool to room tempera-
ture in cold water.
4. Read in a spectrophotometer against reagent blank, at 630 nip, in
%T. Obtain value from calibration curve.
5. If value is too high to read, dilute volumetrically with glacial
acetic acid, read, correct by appropriate dilution factor.
Reagent :
Mix 60 ml. o-toluidine with 1.5 gms thiourea in 1 L volumetric
flask. Dilute to volume with glacial acetic acid, mix well. Transfer
to brown glass bottle, Keeps at least 2 months at room temperature
but must be well stoppered. It should be clear and colorless.
Note :
1. The reaction occurs only with aldohexoses, and will therefore
react with galactose. The color follows 8eers Law in concentra-
tions up to 1000 mg. percent.
2. Solutions must be at room temperature when read. Warm solutions
give falsely high values. Color is stable for 20-30 minutes.
3. o-toluidine will vary from batch to batch, and a new calibration
curve must therefore be made when new stock toluidine is used.
-------
- 50 -
Most commercial o-toluidines are not pure enough, Eastman Kodak
is at present the best available. Thiourea is added to stabilize
the reagent and to prevent turbidity.
References : the present method is a modification of several in the
literature.
1. Nature, Vol. 183, p. 108, 1959.
2. Clin. Chem. Acta, Vol. 7, p.140, 1962.
3. Clin. Chem., Vol. 8, p.215, 1962.
------- |