For Reference
NOT TO BE TAKEN FROM THIS ROOM
Gx UJW*
anrommais
aumainsis
University of Alberta
Digitized by the Internet Archive
in 2018 with funding from
University of Alberta Libraries
https://archive.org/details/dehydrogenationoOOwank
THE UNIVERSITY OF ALBERTA
DEHYDROGENATION OF n-PROPANOL ON AN ALUNDUM CATALYST
BY
SIEGHARD E. WANKE
A THESIS
SUBMITTED TO THE FACULTY OF GRADUATE STUDIES
IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE
DEGREE OF MASTER OF SCIENCE
IN
CHEMICAL ENGINEERING
FACULTY OF ENGINEERING
DEPARTMENT OF CHEMICAL AND PETROLEUM ENGINEERING
EDMONTON, ALBERTA
APRIL, 1966
ABSTRACT
The catalytic vapor phase reactions of n-propanol were
investigated using a sodium hydroxide-treated Alundum catalyst
in a stainless steel fixed-bed flow reactor. The Alundum had
been transformed from a largely acid site dehydration catalyst
to a dehydrogenation catalyst by treatment with sodium hy¬
droxide.
The effects of temperature and contact time on the pro¬
duct distribution were determined over the following ranges
of operating conditions.
Temperature: 300° to 500°C at a constant space velocity
-2
of 1.04 x 10 (gm-moles of n-propanol)/
(hr-gm of catalyst)
-2 -2
Space Velocity: 0.615 x 10 to 19.16 x 10 (gm-moles of
n-propanol) per (hr-gm of catalyst) at a
constant temperature of 463°.
A reaction model, similar to an earlier model described
by Vasudeva, but extended to include several other compounds
identified as products is presented. This model involves,
the dehydrogenation of n-propanol to propionaldehyde followed
by condensation of the aldehyde to form either its aldol or
the ester, n-propylpropionate. These two condensation pro¬
ducts may individually react to form diethyl ketone. The
ester route itself involves two alternative parallel
routes to diethyl ketone, either a direct step or
; . . -i.r : , 1
woll 4 '
i
•f. . fttfj v i J H: s’..:' 3h. . oQl£b\yb-.: B 03
- : -r ■' 'i.i : r.r. - . .• - >. . ■ A" . v.
■ • . ie?c ■ • . ■
, s-k .
\U ■ < . v \ "o) 01 ■ 1-0 ».i to
(3 - • ' 0 ;
' ■ ' .
v : : a.-..; .or v * f j. ffi.; a .JO ••:£ A
e&r; au
■
A-' AV ‘ •• " .-v- . £ A, : 3 T.iJ
■
i - .
. . . ■ ' , ,o
hydrolysis to propionic acid which reacts further to yield
diethyl ketone. Intermediate compounds identified in the
product mixture, as well as experiments with intermediate com¬
pounds as feed, gave evidence that all three proposed routes
contribute to the diethyl ketone formation.
The presence of 2-pentene, the dehydration product of
3-pentanol, indicates that dehydration reactions are also
occurring. The occurrence of thermal decompositions is shown
by the presence of methane, ethane and ethylene in the reaction
products. Evidence was also found that high molecular weight
compounds, believed to be the result of continued aldol con¬
densations of propionaldehyde, are formed. Some evidence is
also presented showing that the stainless steel reactor sur¬
face also acts as a dehydrogenation catalyst.
Overall rates of formation, based on the bulk phase
concentration, for eight of the compounds encountered in the
postulated reaction sequence were determined from the experi¬
ments at 463°C.
:•* . ? • *1 ;-ax c'oirfw ftxa.s. t.-o K , ':u,’ c**.
-
i . . s , . - ' ■ - I ■ ■“
- : . • ■ ... r. J , y p...
. . a j.i , •- • < ■: .-, ■; a ■ iSSv. . , :
u "a.Aavaa.p . ' ■ .. .-x-ii/. ' , ...' a.. -.. "
ax : 'j,; . , a... ax i &oa®a!3:ta s/U £Cf
' - . .
: 5 . . ' . .
.
... . • 5 S
, . ; : .o i' : io ®. . . ■ X " ,• ■ -.'H;
. ~:c ' '■ ' ,3t;■ i:- / ......
ACKNOWLE DGEMENTS
The author is indebted to Dr. I.G. Dalla Lana of
the Department of Chemical and Petroleum Engineering, Uni¬
versity of Alberta, for his guidance and supervision during
the course of this investigation.
The helpful suggestions of Dr. K. Vasudeva in the
early stages of the study are greatly appreciated. Apprecia¬
tion is also expressed to Miss H. Kaethler and my sister,
Ingrid for their patience and assistance in preparing the
manuscript.
The financial support received from the National Re¬
search Council of Canada and the University of Alberta is
gratefully acknowledged.
i
fr-fl <■ s-f s.c-'D 3;C "lie, /-o". xfcn s. •:
TABLE OF CONTENTS
LIST OF FIGURES i
LIST OF TABLES ii
I. INTRODUCTION 1
II. THEORY AND LITERATURE SURVEY 4
A. Alumina as a Catalyst 4
B. Chemical Reactions 5
1. Dehydrogenation and Dehydration of Alcohols 5
2. Condensation Reactions 6
3. Decarboxylation of Esters and Acids 7
4. Thermal Reactions 8
5. Miscellaneous Reactions 11
6. Summary of Chemical Reactions 13
III. PROCESS AND EQUIPMENT 16
A. Equipment
1. Feed System 16
2. Reactor 18
3. Heater 21
4. Temperature and Pressure Control 21
5. Temperature Recording 24
6. Product Collection System 24
7. Glass Reactor v 26
B. Operation of Equipment 26
1. Preparation of Equipment 27
2. Start-Up 27
3. Steady-State Period 29
4. Shut Down 30
C. Raw Materials 31
D. Analysis of Products 33
1. Liquid Products 33
2. Gas Products 36
3. Accuracy of Analysis 38
IV. EXPERIMENTAL PROGRAM AND RESULTS 40
A. Initial Studies 43
1. Preparation of Catalyst 43
2. Activity of the Catalyst 44
3. Stainless Steel as a Catalyst 44
TT^-' -r jo t ass ■ . o;/'
. 3 . :
. - ' .■ ■ ' L
■!. 3 'V. ' ■ 3 '■
B. Influence of Temperature and Space Velocity on
Catalytic Activity 46
1. Effect of Temperature 46
2. Effect of Space Velocity 46
C. Special Runs 57
D. Poisoning of Catalyst 57
V. DISCUSSION AND INTERPRETATION OF RESULTS 64
A. Chemical Reaction Sequence 64
1. Model of Major Chemical Reaction Sequence 65
2. Discussion of Reaction Model 67
3. Aldol Versus Ester Route 70
B. Reaction Rates 76
1. Overall Rates of Formation 76
2. Factors Influencing Reaction Rates 77
3. Methods of Correlating Rate Data 81
4. Correlation of Rate Data 84
C. Yields 89
VI. CONCLUSIONS 94
VII. RECOMMENDATIONS 96
BIBLIOGRAPHY 98
APPENDIX I - THERMOCOUPLE EVALUATION 1-1
APPENDIX II - SAMPLE TEMPERATURE PROFILES II-l
APPENDIX III - PRODUCT ANALYSIS III-l
APPENDIX IV - OPERATING CONDITIONS AND PRODUCT
ANALYSIS FOR EXPERIMENTAL RUNS IV-1
APPENDIX V - OVERALL RATES OF FORMATION DATA V-l
.>. .. .a .v ; *fc ..
. -■)'} . : ■ ' •,
’
1
LIST OF FIGURES
Figure Page
1 Diagrammatic Layout of Equipment 17
2 Vertical Cross-Section of Reactor 19
3 Location of Thermocouples 20
4 Diagrammatic Cross-Section of Heater and
Eutectic Bath Showing Location of Reactor 22
5 Effect of Temperature on n-Propanol
Conversion and Diethyl Ketone Yield 50
6 Effect of Temperature on Liquid Product
Distribution 50
7 Effect of Temperature on Gaseous Product
Distribution 51
8 Effect of Space Velocity on n-Propanol
Conversion and Diethyl Ketone Yield 55
9 Effect of Space Velocity on Liquid
Product Distribution 55
10 Effect of Space Velocity on Gaseous
Product Distribution 56
11 Effect of Time on Catalyst Activity 63
12 Reaction Model 66
13 Overall Rates of Formation 78
14 Test for First-Order Reaction for Decompo¬
sition of n-Propanol 86
15 Temperature Dependency of the Rate of
Decomposition for n-Propanol 87
\i-.'■ -V. n :.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
- li -
LIST OF TABLES
Page
Properties of Catalyst 32
Performance of Base-Exchanged Catalyst 45
Effect of Stainless Steel and Glass on
the Decomposition of n-Propanol 47
Effect of Temperature on Catalytic Activity 48
Effect of Temperature on Product
Distribution 49
Effect of Space Velocity on Catalytic
Activity 53
Effect of Space Velocity on Product
Distribution 54
Comparison of Catalytic Activity of Base-
Exchanged Alundum to Base-Exchanged Chromia
on Alundum 58
Operating Conditions and Product
Distribution for Runs with Special Feeds 59
Life of Catalyst 61
Effect of Time on Catalytic Activity 62
Comparison of Calculated to Experimental
Yields of Diethyl Ketone and Hydrogen 72
Effect of Water-Gas Shift Reaction on
Carbon Monoxide and Carbon Dioxide Yields 75
Effect of Temperature on Rate Constant
for Reactions of n-Propanol 88
Comparison of Diethyl Ketone Yields 91
"o f •„ : v ;■ • V ; :
'
1
I. INTRODUCTION
The catalytic dehydrogenations of primary alcohols to
aldehydes and secondary alcohols to ketones are well known
industrial processes, but the dehydrogenation of primary al¬
cohols to ketones is not a commercial nor fully understood
operation. Ipatieff(10) was one of the earliest to report
the dehydrogenation of ethanol to acetone and later, Komarewsky
and Coley(13) investigated the dehydrogenation of a large num¬
ber of primary alcohols. They reported that a primary alco¬
hol containing n carbon atoms can readily be converted to a
symmetrical ketone containing 2n-l carbon atoms using precipi¬
tated chromia as a catalyst. A 48% yield of diethyl ketone
was reported for the conversion of n-propanol to diethyl
ketone.
On this basis, an investigation of the industrial
possibilities of this process was undertaken at the University
of Alberta by Hansen(5) in 1959. Hansen studied the dehydro¬
genation of n-propanol over an alumina supported chromia
catalyst in a fluidized bed. A fluidized bed was used to
overcome possible temperature gradients due to the endothermic
nature of the reactions. A supported catalyst was used since
the precipitated chromia did not have the physical strength to
withstand fluidization. Hansen obtained a diethyl ketone
yield of only 8.7% and concluded that the alumina supported
chromia catalyst was unsuitable for converting n-propanol to
■ •;. ' . • ' . 1 V
■
I .1 ' : gj/ : . r „ .
■••• ' ' " if/ - .tti ■. .1 ' , if-.;- , L ; , 'I
■ tr-. ;■ ' 1 /-■/: a/ , - ■■ *
' -
* t •
'
2
diethyl ketone.
Vasudeva(31) continued this work using a fixed bed
reactor to investigate the effect of the support. He found
that the alumina was dehydrating, converting n-propanol to
n-propyl ether and water at low temperatures and to propylene
and water at higher temperatures. When the alumina was treated
with sodium hydroxide, this dehydration was suppressed and de¬
hydrogenation started to occur. With an alumina-chromia
sodium hydroxide catalyst, Vasudeva obtained a 15.2% diethyl
ketone yield.
Subsequent work by Vasudeva(32) was concentrated on
investigating the chemistry of this system, since a disagree¬
ment in the reaction sequences was reported between those of
Ipatieff(10) and of Komarewsky(13). Besides identifying
several more products, which helped to verify his proposed
reaction sequence, Vasudeva also obtained a supported chromia
catalyst which gave a diethyl ketone yield of 48%.
In this work, Vasudeva(32), reported obtaining a
42.2% diethyl ketone yield using a sodium hydroxide treated
alumina catalyst without chromia. This catalyst, if capable
of giving yields comparable to an impregnated chromia catalyst,
has the advantages of much lower cost and easier preparation.
For this reason, it was decided to evaluate this base-treated
catalyst for the production of diethyl ketone from n-propanol.
This work was therefore undertaken to investigate the
■
'
■ • \ . i;; „•
• ' ■
3
reactions of n-propanol in the presence of a sodium hydroxide
treated alumina catalyst. This investigation consisted of
finding the influence of space velocity and temperature on
the activity as well as the selectivity of the catalyst.
In order to evaluate the selectivity of the catalyst
a thorough knowledge of the chemistry of the system, as well
as a reliable analysis of the products, is required. As a
result an effort was made to improve on the reaction sequence
postulated by Vasudeva and also to improve the product analysis.
■ 1-'■ -V :
4
II. THEORY AND LITERATURE SURVEY
Vasudeva(32) presents a review of the catalytic
activity of alumina and chromia catalysts. He also lists the
possible reactions of n-propanol in the presence of these
catalysts. The following sections summarize the catalytic
properties of alumina and give a detailed description of the
possible vapor phase reactions of n-propanol in the presence
of a sodium hydroxide treated alumina catalyst.
A. Alumina as a Catalyst
Alumina can exist in seven different crystalline struc
tures and its catalytic activity, to some degree, depends on
which structure is present. In general, it may be said that
alumina catalyses reactions such as polymerization, isomeriza¬
tion, cracking and alkylation of hydrocarbons, and dehydra¬
tion of alcohols(32). Since these reactions are also acid
catalysed, the catalytic activity of alumina is usually at¬
tributed to "acid sites" on the surface of the alumina(20)
Upon treatment with sodium hydroxide, the acid centers
on the alumina should be partially or wholly destroyed. This
was born out in studies by Vasudeva(32), who showed that the
dehydrating activity of alumina towards n-propanol was com¬
pletely destroyed when the alumina was treated with 2N NaOH
solution. The interesting behavior was not the suppression of
dehydration, but rather that the base exchanged alumina now
became a dehydrogenating catalyst.
'
fi . .C ■ ' •
<1 -• -8 3 tc ,tO®8 loi © ;1T , • • J:
■ ■ ' ■ . r .
.■ - • J ' :• ■ e"s *3 /:• . ;■ Hi"'
• : 0
snod- brrs A r. jj-
sloriojaifl
■ f -
‘ ■ : ■ «-■' ■■■ - ' " • ■ --VV r. 9 / 1 ; j-,rl . V . -■
5
B. Chemical Reactions
1. Dehydrogenation and Dehydration of Alcohols
The dehydrogenation of primary and secondary alcohols
to aldehydes and ketones, respectively, are well known chemical
reactions. The aldehydes formed in this way usually undergo
further reactions unless a selective catalyst, such as copper,
is used. The ketones, on the other hand, are much more stable
towards secondary decomposition. A detailed discussion of the
dehydrogenation mechanism, using ethanol, is given by Church
and Joshi(2).
A parallel reaction to dehydrogenation is the dehydra¬
tion of alcohols to form the olefin and water. Many metals
and their oxides, such as sodium, magnesium, aluminum, nickel,
iron or zinc, catalyse both the dehydration and dehydrogena¬
tion of alcohols(15).
Ipatieff(10) was one of the first to study the decom¬
position of alcohols with an alumina catalyst in an iron tube.
When using ethanol as feed he obtained, among other products,
water, ethylene, hydrogen and acetaldehyde. To explain these
products he suggested the simultaneous dehydrogenation and
dehydration.
c 2 h 5 oh -ch 3 cho + h 2 .(1)
C o H c 0H
2 5
* C 2 H 4
+ h 2 o
( 2 )
...
rjjg - J '
V. ,;-fo 1 o r:'^i j.:,:/" . ,;: ; T
■ »q.8i
it ■' ■.■■ "
: ' " , '
1 '■ ■ ■' ’ a.. :• ■' . ,
, (a.) i&Cr. bnfl
ri:--; ■ od fra : vos •: ic i ' nit [ f
■ mi
6
Here it was thought that the dehydration was caused
by the alumina and the dehydrogenation by the iron walls of
the reactor.
2. Condensation Reactions
There are two possible routes by which aldehyde con¬
densation can occur, the aldol condensation and Tischenko con¬
densation (13,26). The aldol condensation is catalysed by both
acids and bases, while the Tischenko reaction is catalysed by
oxides of such metals as zinc, chromia, cobalt or aluminum.
Writing these reactions for propionaldehyde we obtain
2C 2 H 5 CHO -
-► C~H c CH (OH) CH (CH^) CHO
(3)
2C 2 H 5 CHO -
-* c 2 h 5 cooc 3 h 7
(4)
The aldol formed by reaction 3 is fairly unstable and
may undergo dehydration or decarbonylation,
C 2 H 5 CH(OH)CH(CH 3 )CHO - C 2 H 5 CH = c ( ch 3 ) ch0 + H 2°
c 2 h 5 ch(oh)ch(ch 3 )cho -c 2 h 5 ch(oh)c 2 h 5 + CO (6)
The dehydration of the aldol to the unsaturated alde¬
hyde and water is an acid catalysed reaction(3). Komarewsky
and Coley(13) through isotrope trace studies and from the
amount of carbon monoxide formed, concluded that aliphatic
aldehydes in the presence of a chromia catalyst undergo aldol
condensation followed by decarbonylation to give the secondary
f 5)
■ " " ' ' ' o'i : ■ .. y t
■ nc >
-
‘ • ■ -ftiT ,• ■ / . ,1 :• ■;
..
7
alcohol. In the presence of this catalyst the secondary al¬
cohol dehydrogenates to form the symmetrical ketone,
c 2 h 5 ch(oh)c 2 h 5 -► c 2 h 5 coc 2 h 5 + h 2 (7)
Ipatieff(10), on the other hand, postulated that the
Tischenko condensation, due to the ester found in the product
and the large amount of carbon dioxide formed, was the major
route to ketone formation when he studied the decomposition of
ethanol in an iron reactor using alumina as a catalyst.
3. Decarboxylation of Esters and Acids
Both acids and esters in the presence of a suitable
catalyst will react to form ketones, olefins, carbon dioxide,
water and alcohol. There are several routes by which the
ester can undergo this decomporition, depending on whether or
not water is present(ll). Kagan et al(12) investigated the
reactions of ethyl acetate in the presence of copper, zinc,
and chromium oxides, and combinations of these, both with and
without water in the ethyl acetate. They showed that even in
the absence of water the ester can decompose to the ketone,
the postulated reaction being
2c 2 h 5 cooc 3 h 7 -»- c 2 h 5 coc 2 h 5 + c 3 h 6 + co 2 + c 3 h ? oh (8)
If water is present the ester first hydrolyses to the
acid which in turn reacts to give the ketone.
c 2 h 5 cooc 3 h 7 + h 2 o
* C o H c C00H + C^H_OH
2 5 3 7
(9)
.
..
arid no] - ■■ .. r .• ql
' ' ... ■ ' ■ 1- ■■■"„ : v -: r&r>
... ' :ia
.
■ ■ -V... •■'.{, I- . :
■ X) : t'-St: - .}
0
2C 2 H 5 COOH -* C 2 H 5 COC 2 H 5 + C0 2 + II 2 0 (10)
If the hydrolysis of reaction 9 occurs in a basic
media, the equilibrium is far to the right(18). Besides the
well known vapour phase reaction of aliphatic acids over an
oxide catalyst to yield ketones(19), the acid can also de-
carboxylate to give the corresponding alkane.
c 2 h 5 coqh ——-* c 2 h 6 + co 2 (11)
For the comparison, the overall decompositions of
ester to ketones with and without water are
2c 2 h 5 cooc 3 h ? ——► c 2 h 5 coc 2 h 5 + c 3 h 6 + co 2 + C 3 H ? OH (8)
2C 2 H 5 COOC 3 H ? + 2H 2 0 -->C 2 H 5 COC 2 H 5 + C0 2 + H 2 0 + 2C 3 H 7 OH (12)
It can be seen that these two reactions become identi¬
cal if the alcohol that is formed dehydrates to the olefin,
then the overall reaction for both routes becomes
2c 2 h 5 cooc 3 h ? --► c 2 h 5 coc 2 h 5 + co 2 + c 3 h 6 + C 3 H ? OH ( 13 )
4. Thermal Reactions
Hurd(8) and Rice(23) did extensive studies on the
thermal decomposition of organic compounds, and their books
give an excellent review of this subject.
Rice(23) reports that the mechanism of decomposition
of primary alcohols is a direct separation of molecular
hydrogen from the alcohol to form the corresponding aldehyde.
'• .• - v..» .
>, v- ■ »■ ■■ . ■ . ■■ I'. ' v ./ ;• ■ , j S' ; :'■. >ro
. . -.. — . . ...i
.: •: K-'r-. : •'
V ; V. Ol i.: L : ■ : *•' - - ; "< . t J <j~
;■ ■ - .. 1 :■ ■ U : ‘
■, r - ■ . .; ■:/. r,' • .; :
■jjf.'cU. iljod OSt TtC j " 3"£ ’ s' J /:'• ‘.i OV-O-
. VO ••'SH J V J>.±Ti'
j '
le>,po . ■ q ■ ■ ' " ' ■ ■
.
. - 'j r~.r . fe: ; xo>' '"a.
■■■■•■ : : ■. H '' x •; ' ■ ■;:
„ : - /. . j' C f '• ‘ y •
9
RCH 2 OH -* RCHO + H 2 (14)
Hinshelwood and Thompson(7) studied the decomposition
of propionaldehyde in the 450° to 600°C temperature range and
found that the decomposition
C 2 H 5 CHO -* C 2 H 6 + CO (15)
was the predominating reaction. At 500°C the products formed
from the decomposition of propionaldehyde consisted of 51%
carbon monoxide, 26.5% ethane, 18% methane, 2% carbon dioxide
and 2% ethylene. The effect of increasing the temperature
was to increase the methane content. They explain the forma¬
tion of methane by a simultaneous decomposition of the pro¬
pionaldehyde into methane, ethane and carbon monoxide rather
than by the pyrolysis of ethane to ethylene and methane since
the ethylene content of the product was always low.
The pyrolysis of paraffins is a well-studied field
and the decomposition products of propane would be hydrogen,
ethylene, propylene, and methane. Rice(23) reports that at
300°C the reaction of propane would be
C 3 H 8 -* C 2 H 4 + CH 4 (16)
while at 1000°C, the reaction
C 3 H 8 -- C 3 H 6 + H 2 (17)
would be the important one.
' ' ' ' .
: " : . - ■ ■ : : ;
• ^ V_ ' v ' ... .
:• . r .. rv-'
• - : '• ' "■'->. ■.. . . . ,
10
The pyrolysis of olefinic hydrocarbons, such as propy¬
lene, are much more involved than that for the corresponding
saturated hydrocarbon because of the great variety of pro¬
ducts that can be formed from polymerization and condensation
reactions. Hurd and Meinert(9) showed that propylene was
stable up to 600°C in a pyrex apparatus, but when the pyroly¬
sis was carried out in a monel metal tube, decomposition was
almost complete at 375°C. The decomposition was accompanied
by the deposition of elemental carbon on the reactor walls
which they believed catalysed the decomposition. Schneider
and Frolich(27) also noticed the formation of compounds such
as butanes and butenes from the pyrolysis of propane and pro¬
pylene. All these products can readily be explained by a free
radical mechanism.
Hinshelwood and Hutchinson(6) studied the thermal de¬
composition of acetone and found the main products to be
carbon monoxide, ethane, ethylene and hydrogen with traces of
carbon dioxide and higher condensed products. For larger
ketones the products become much more complex and Rice(23)
outlines some of the products that are formed from the de¬
composition of larger ketones such as methyl ethyl and diethyl
ketone.
The thermal reactions for reaction mixtures contain¬
ing a large number of compounds, such as aldehydes, ketones,
olefins and paraffins, become very complex because of the
large amount of products possible from the interaction of the
,'r.L . ' O " ;\ls XU/X.X/q V..'.'."
t .. ■ •• " - .3, : -■;<■ r ■■ y ... ' , ?• 3 y s ,a v >:.i , : “ssi
■ '■ y ... •' r. y.yyyyyy nr ' irr y . v. ....
■. ■ ' ■ k .
ta '
■ K; ; ^.?• . •- h ' aJ: i:x« ; s s ...■ :? ,<:&
: <- ;. 3... . ; i - ":|£ 'y> ed.t ,>.1
.
' -ft. 3..- . ' :h :;:t yziy.ir ' , -
- , ;.. . •: 3.,.:;/, .1 ■
vavbaoo bn* y-yyy c. i.b
■ ' • . 1 . ? . r, . . • /
' 1 3; '7.(3/ . V ■ ‘Wll- . - ■ •
V--,-" .. it-A y.ry'x.,\i ix. ,y:.-yzyy ? .pv>.3
3 '■ '■ ' y.y. oo "tr • a iu i -. ■■■■.;:% %y: ■;
i
11
different compounds.
5. Miscellaneous Reactions
Besides the reactions already outlined there are sever¬
al other reactions that the products encountered in the dehydro¬
genation of n-propanol can undergo that do not fit into a gen¬
eral classification.
Noller(19) reports the reaction of an olefin with an
aldehyde to give a ketone,
c 2 h 5 cho + ch 2 = chch 3 -► c 2 h 5 coc 3 h ? (18)
This reaction is a free radical reaction and requires the
formation of the CH 3 CH 2 C = 0 radical.
Many workers have observed the formation of ethane in
the dehydrogenation of ethanol over a metallic oxide catalyst.
Engelder(14) explained this by the hydrogenation of ethylene
which is formed by the dehydration of the ethanol.
C 2 H 4 + H 2 -- C 2 H 6 (19)
Adkins(28), however, suggested that the ethane is formed by
the reaction,
2C 3 H ? OH _* C 2 H 6 + CH 3 CHO + H 2 0 (20)
Neither reaction is satisfactory. Engelder's pro¬
posed hydrogenation is unlikely because the reaction condi¬
tions basically favor dehydrogenation, and Adkins' reaction
on the other hand, is only a balanced chemical equation which
. : - 1 'Jtv
■ .
Xi- " . L» f IOv ■ l:.J. . \
s''x — ■' ! *
12
contains the observed products without giving a mechanism by
which this can occur.
To explain the large amount of carbon dioxide formed
during the dehydrogenation of normal alcohols, Lazier and
Adkins(14) proposed the formation of ketenes as an intermedi¬
ate. The ketene would react with an aldehyde to produce the
olefin and carbon dioxide.
C 2 H 5 CHO -CH 3 CH == C = 0 + H 2 (21)
CH-CH = C = 0 + C„H c CHO -*■ C 0 H c CH = CHCH- + C0~ (22)
3 2 5 2 5 3 2
Besides the reaction of ketene with aldehydes, ketenes can also
react with water to form the acid or with hydrogen to regen¬
erate the aldehyde. Also, ketenes as well as aldehydes readily
undergo polymerization as outlined by Bevington(1).
Vasudeva(32) reports evidence for the reaction of di¬
ethyl ketone to yield methyl ethyl and ethyl isopropyl ketones,
2C o H t -C0C o H t - -CH o C0C o H c + C o H c C0CH (CH_ ) „ (23)
25 25 3 25 25 32
a possible mechanism for this reaction involves an aldol type
condensation of the ketone followed by an alkyl rearrangement
and a bond cleavage to yield the observed products.
The water-gas shift reaction is
CO + h 2 o ... co 2 + Il 2 (24)
This disproportionation reaction is encountered whenever a
gas mixture contains the compounds in equation 24 in quantities
"
, o :.a& n.
rs-tf , j- .. ■ Y“’- ■ - ® r t '
,v... .. 3 fS i< !VXdq f -
•: J - oo - ■ f. .
13
other than equilibrium concentrations. Metals and metallic
oxides catalyse this reaction. Reaction 24 is reversible
and the equilibrium is very temperature dependent.
Values of the equilibrium constant, K , are 11.7, 7.3
P
and 4.9 at 400°, 450°, and 500°C, respectively(21).
Where K is defined as
P
[CO,][H,J
K = ---—
P [CO][h 2 o]
6. Summary of Chemical Reactions
The following list of reactions covers the main chemical
reactions which n-propanol and its subsequent reaction products
may be expected to undergo in the presence of a sodium hydrox¬
ide promoted alumina catalyst.
a) dehydrogenation of n-propanol
C 3 H ? OH -*■ C 2 H 5 CHO + H 2 (1)
b) dehydration of n-propanol
C 3 H ? OH -•* C 3 H 6 + H 2 0 (2)
c) aldol condensation of propionaldehyde
2C 2 H 5 CHO -*■ C 2 H 5 CH(OH)CH(CH 3 )CHO (3)
d) dehydration of the aldol
* c 2 h 5 =c(ch 3 ) CHO + h 2 o
C 2 H 5 CH(OH)CH(CH 3 )CHO
(5)
. . >■' i'' > : . •'.* ' ' ■ - -■
.. /: j: ,.y ■ ; ■ "■>
%> .
■
.. >1. /.; * r, ;j " - * . .■ r - m ;
..
■
14
e) decarbonylation of the aldol
C 2 H 5 CH(OH)CH(CH 3 )CHO -► C 2 H 5 CH(OH)C 2 H 5 + CO (6)
f) dehydrogenation of 3-pentanol
c 2 h 5 ch(oh)c 2 h 5 -► c 2 h 5 coc 2 h 5 + h 2 (7)
g) dehydration of 3-pentanol
C 2 H 5 CH(OH) C 2 H 5 -» C 2 H 5 CH=CHCH 3 + H 2 0 (25)
h) Tischenko condensation of propionaldehyde
2C 2 H 5 CHO ————► C 2 H 5 COOC 3 H 7 (4)
i) hydrolysis of the ester
c 2 h 5 cooc 3 h ? + h 2 o—- + c 2 h 5 cooh + c 3 h 7 oh (9)
j) decomposition of the ester
2 c 2 h 5 cooc 3 h 7 - c 2 h 5 coc 2 h 5 + c 3 h 6 + co 2 + c 3 h ? oh (8)
k) reaction of propionic acid
2 c 2 h 5 cooh -*. c 2 h 5 coc 2 h 5 + co 2 + h 2 o ( 10 )
l) water-gas shift reaction
CO + H 2 0 - - C0 2 + H 2 (24)
Besides these reactions there occur a number of side
and thermal reactions to produce methane, ethane, carbon
iSi x bn:
15
dioxide, carbon monoxide, hydrogen and a host of higher con¬
densation products.
16
III. PROCESS AND EQUIPMENT
A. Equipment
A schematic diagram of the equipment used in the de¬
hydrogenation study of n-propanol is given in Figure 1. The
equipment may be divided into the following sections: feed
system, reactor, heater, temperature and pressure controls,
and product collection system. Each of these sections will
now be discussed in detail.
1. Feed System
The feed system was constructed so that either gas or
liquid feed could be introduced to the reactor but not simul¬
taneously. The only gas feed used in this work was nitrogen
and it was supplied by a pressurized nitrogen cylinder. The
pressure of the nitrogen supply was regulated at the cylinder
by a gas pressure regulator, and the flow rate of nitrogen was
controlled with the gas rotameter valve.
The liquid feed, stored, in a calibrated reservoir of
160 ml capacity, could be pressurized with nitrogen by opening
and closing the appropriate valves as shown in Figure 1. The
pressure in the liquid reservoir during operation was main¬
tained at 40 psig by adjusting the pressure regulator on the
nitrogen cylinder. The liquid flow was controlled with the
liquid rotameter which was a Matheson 602 stainless steel rota¬
meter. A second auxiliary feed tank attached to the main
.
■■■■■ : -■ 7,0; r>|
Vent
17
Figure 1: Diagrammatic Layout of Equipment
18
liquid reservoir could be used if a larger amount of liquid
feed was required. This reservoir was not used since the 160
ml capacity of the main reservoir was sufficient.
All the lines in the liquid feed system were made of
1/4-inch, 316 stainless steel tubing, while 1/4-inch copper
tubing was used as nitrogen lines.
2. Reactor
The reactor was fabricated from a 12-inch length of
3/4-inch nominal 316 stainless steel pipe. Figure 2 shows a
vertical cross-section of the reactor giving dimensions and
construction details. Eight thermocouples were attached to
the reactor, their location being shown in Figure 3. As in¬
dicated in Figure 3, two thermocouples were joined to the
outer wall of the reactor, five were radially connected to the
reactor so that each tip extended to the centerline of the
reactor, the remaining thermocouple being placed at the en¬
trance to the reactor. The five thermocouples that were in¬
side the reactor were bent 90° one-half inch from the tip so
that the ends were pointing downwards at the centerline of
the reactor. This was done to minimize thermocouple errors
due to conduction (see Appendix I).
The thermocouples used were ceramic-insulated, metal-
sheathed 1/16 inch O.D. iron-constantan thermocouples pur¬
chased from Thermo Electric Company (Catalogue No. 5J0411E).
'
f ■ ’ ■-$ '^O* i ^ ic ri : . ,.,^,v
‘ ' ■
• V - n '" h ~“'- - ni ^ ! --'^.00 ; -..,.:V:,., -OC'O.6*,.' Cul'r>
r * ' n:C '' i: ' '■■ ; £«W */ii ,:,VD-Xi H rf o f : &■■■
■ *® 1 i A* JA.fna, , lc v..-i
" " - i/OMr ui * »* '•‘AiaJjtt oc iMioS ajsv ;;c
'Jcico ;huyy :y utr.fr
- 19 -
FIG 2 VERTICAL CROSS-SECTION OF REACTOR
Key to Figure 2
TCI, TC2 and TC8 - 1/16" O.D. iron constantan thermocouples
(Thermoelectric Company Cat. No. 5J0411E)
1 - 1/8" to 1/16" reducing union (Swagelok Cat. No.
200-6-1-316)
2 - 1/8" stainless steel tubing
3 - 1/4" to 1/8" reducing union (Swagelok Cat. No.
400-6-2-316)
4 - 1/4" stainless steel tubing - 18 gauge
5 - 1/4" stainless steel tubing - 22 gauge
6 - 1/4" union tee (Swagelok Cat. No. 400-3-316)
7 - 1/16" male connector (Swagelok Cat. No. 100-1-1-316)
8 3/4" Nominal stainless steel pipe
9 - 200 mesh stainless steel screen
10 - Press fit ring 1/2" I.D. and 5/8" O.D.
11 - 1-3/32" fitting (Swagelok Cat. No. 1210-10-316)
12 - Stainless steel cover for heating bath
• •■■•pi- i
■
?■■■: ;r ( •• ; . : .
. :
■ ■ ' • ■■ - ' .• :i . 0 : '
■ ■ ; ■
■
20
Figure 3: Location of Thermocouples
21
3 0 Heater
The reactor was heated by immersing it in a cerrobase
bath, i.e. an eutectic composed of 55.5% bismuth and 44.5% lead
by weight, which melts at 130°C(20). The eutectic was heated
by a 2500 watt calrod heater placed as a helical coil on the
inside wall of the 6.5 inch diameter, 18 inch high vessel con¬
taining the cerrobase and the reactor. A schematic cross
section of this unit showing the location of the reactor in
the metal bath is given in Figure 4. The whole container was
surrounded by 4.5 inches of asbestos insulation. The bath was
also equipped with a stirrer driven by a variable speed 1/4 hp
electric motor. The stirrer was installed to provide better
heat transfer to the reactor, but by experimentation it was
found that the thermal conductivity of the cerrobase was suf¬
ficiently high that no appreciable temperature gradients
existed in the bath even with the stirrer shut off. For this
reason, the stirrer was not used in this experimental work.
In order to vaporize and preheat the feed 8 feet of
1/4 inch stainless steel tubing were immersed in the heating
bath and connected to the reactor inlet. At n-propanol feed
rates higher than 70 gm/hr., the preheater was not sufficient
to heat the vapor to the bath temperature.
4. Temperature and Pressure Control
The temperature in the eutectic bath was controlled by
a Foxboro Model 64100 FO temperature recorder-controller. A
. t-i, ,r „L ■oJo&o't I-
' ' ■, ■ • ■ ■
Inle
22
Figure 4: Cross-Section of Heater and Eutectic
Bath Showing Location of Reactor
(Thermocouples not shown. See Figure 3)
23
Foxboro type 693 EMF converter equipped with a 1/4 inch iron-
constantan thermocouple supplied the input to the controller
which regulated the current to the calrod heater. Very good
temperature control was obtained in this way and the bath tem¬
perature did not fluctuate more than 1°C once steady tempera¬
ture in the bath was obtained. The current drawn by the calrod
heater could be read by an ammeter attached to the circuit.
For n-propanol feed rates of 50 gm/hr., a current of approxi¬
mately 3 amp was required to maintain a constant temperature
in the reactor and the heating bath. The temperature range
over which the controller will regulate the temperature de¬
pends on the range unit in the EMF converter. A 300° to 500°C
range unit was sufficient for this work.
The pressure in the reactor was regulated by adjusting
the throttling valve so that the reading on the 5 foot mer¬
cury manometer, attached to the inlet line of the reactor,
corresponded to the required pressure. All experimental runs
were carried out at a reactor pressure of approximately 20 psig.
The pressure control obtained by the manual adjust¬
ment of the throttling valve was not entirely satisfactory,
since pressure fluctuations up to 1 psi were encountered
during the operation of the reactor.
'
■ O'. • £?, --
■
*' ' : ■-* ‘ '"9. '**$ .-'Si Xoi . ’ '' : c "
' ; S ' •' ‘ ' -•Xiw -I... ■) o ' ' v„
' 7 ' : ' ; ■ --'.a • S3 3SOb:'.3l.B . ;. ::W
24
5. Temperature Recording
The eight thermocouples, TC 1 to 8, from the reactor
were connected to a Speedomax Type G Leeds and Northrup' 12
point variable range variable zero temperature recorder. The
time required to complete the 12 point cycle was 48 seconds.
The range normally set on this recorder was 300° to 500°C.
This range was checked, and adjusted if necessary, prior to
every run by feeding a voltage, equivalent to that produced by
an iron-constantan thermocouple at 300°C and at 500°C, into
the recorder.
Another thermocouple, TC9, located on the reactor exit
line before the product collecting system, was connected to a
separate Leeds and Northrup potentiometer since this tempera¬
ture was below 300°C and therefore out of the set range of the
temperature recorder. An ice-water cold junction was used for
all thermocouples other than the one attached to the tempera¬
ture controller since the controller was equipped with a
cold junction compensator.
6. Product Collection System
To prevent condensation of the vapor products in the
line between the reactor exit and the cold traps, an insulated
heating coil was wrapped around this line. This coil had a
resistance of 17 ohms and approximately 55 volts were required
to maintain a temperature of 200°C at thermocouple TC9.
The product collection system was constructed of glass
..
.
"" ^ ■. - I?- ■ . '.'j £>*?■' • . , ■' • , -
V ’ "
, - ■ 'i)l' --W .
25
and the transition from stainless steel tubing to glass was
achieved by an 18/7 steel ball-glass socket joint immediately
after the throttling valve shown in Figure 1.
During start-up and shut-down, the products were vented
through the 3-way stopcock. At steady state, constant tempera¬
ture, reactor pressure and feed rate, the products were passed
through a cold water condenser, which was only put into opera¬
tion when n-propanol feed rates were higher than 50 gm per hr.,
and then into two cold traps. The first cold trap was immersed
in an ice-water mixture while the second one was maintained at
-30°c by means of a dry ice-CCl slurry. The two collecting
4
vessels were of identical construction except for size, the
first having a 150 ml capacity and the second a 40 ml capacity.
The traps were built so that they could easily be removed
from the rest of the apparatus and weighed without having to
transfer the liquid from the containers, thereby avoiding any
loss of liquid product. At least 90% of the vapors were con¬
densed in the first trap.
The noncondensable gases passed from the second cold
trap to either the wet test meter or the gas sampler. The wet
test meter used was of Precision make with a 0.1 cu.ft. per
revolution capacity. The continuous gas sampler consisted of
a glass carboy filled with a saturated sodium chloride solution
over which the gas sample was collected by withdrawing some of
the solution from the bottom of the carboy. The wet test
meter and the sampler were attached in parallel. The accuracy
" / ■.. : ■ ■ ■. -
" ' • ■■■■■■. . . ■ . ' ■' . . ’ ■■ se ■ ■ •
6 • rjj. V. |j ,1:1. • :„J ; >
'
'
• •- ' : ‘ ■■ ...i, ... ■: ■!. , • ",.... : ■ -
' . ' i . :
.
■ ■* ' J ' /' ' ' ", , ■ . ;.i ' ;
• ' : ' ; , : , .
,
26
of the wet test meter at flow rates of 0.05 cu.ft. to 1.0 cu.
ft. per hour was checked and found to be within 13% of the
actual gas flow.
7. Glass Reactor
Some of the experimental work, for comparison with
Vasudeva's(32) work and preliminary catalyst evaluation, was
carried out in a glass reactor. The reactor was similar to
that described by Vasudeva(32). The reactor consisted of a
45 cm long 22 mm O.D. pyrex tube. The catalyst bed was
supported by a porous quartz plate. The feed was fed by
gravity from a constant-heat burette and the liquid product
collected in a cold trap immersed in a dry ice-carbontetra-
chloride bath. The total gaseous product was collected in a
calibrated glass carboy.
The reactor was heated by an electric furnace surround¬
ing the reactor. The temperatures were measured by three
chromel-alumal thermocouples, two located in the catalyst bed
and one inside the furnace adjacent to the outside wall of
the reactor. The input to the furnace was regulated by a
powerstat.
B. Operation of Equipment
An outline of the procedure followed in obtaining the
experimental results is presented in this section. A de¬
tailed discussion of the preparation of equipment, start-up.
'
■ i
; = ■ M' ^ ' a ..idi
' £ • - .•■■.•'■ yJ i ■'.•■;/ ' ,
«
'• - ■ ' •
' ■ • - ■' •;
' . .
" ■ t .'.j. : '
■'! . .. ' : f , 5s:rs:. : •; e .. .:;y ...
:
• . • v-'S% ' ..
: - - r ' ; ' •, ; ■■■ ■ S
■ ■ 3v r ... - •/ .
: . : i f
.
' . ■ . ■
- . < , . ; . . . ..
27
steady-state period and shut-down phases of the operation is
given.
1. Preparation of Equipment
Before any experimental runs could be made, the reactor
had to be charged with catalyst. This presented certain dif¬
ficulties since the reactor became coated with cerrobase once
it had been immersed in the bath, making it difficult to remove
the large swagelock fitting at the bottom of the reactor through
which the catalyst was charged. After the catalyst had been
placed into the reactor and all the fittings tightened, the
reactor was pressurized to 40 psig and immersed in water to
assure that it was leak-proof. As an extra precaution, the
pressure in the reactor was always maintained at 20 psig while
being immersed in the molten metal bath. This pressure of 20
psig was sufficient to stop any molten metal from leaking into
the reactor even if a leak should develop after immersion.
Due to the difficulty in changing the catalyst charge
in this reactor, only two different catalyst charges were used
in this work.
2. Start-up
Before the calrod heater was turned on, the reactor
was pressurized to 20 psig by nitrogen, and the nitrogen flow
through the reactor regulated at approximately 500 ml per
minute at standard pressure and temperature, by means of the
throtteling valve. During the initial heating period, the
■
28
temperature controller was set on manual control and adjusted
so that the heater drew the full current of 11 amps at 220
volts. Once the cerrobase bath temperature reached 250°C, the
controller was set on automatic control. A proportional band
setting of 20% with a 3 minute reset was used on the automatic
temperature controller to give a minimum of cyling or overshoot¬
ing. Approximately two hours of heating time were required to
achieve the set temperature in the bath. During this time, the
range on the temperature recorder was checked, and the cold
traps prepared and installed. When steady state temperature
conditions in the reactor had been obtained, the nitrogen flow
was shut off, the feed system pressurized to 40 psig with
nitrogen, and the liquid feed to the reactor adjusted by set¬
ting the flow rate desired on the calibrated liquid rotameter.
An additional 1/2 hour was required after the liquid feed
was started to obtain steady state temperature, pressure, and
flow conditions in the reaction system. During this time,
the heating coil on the exit line was turned on and the volt¬
age adjusted to obtain a temperature of 200°C at the end of
the line. The product obtained during the start-up period
was discarded by venting it through the 3-way stopcock. Once
steady state conditions had been achieved, the product col¬
lection was started.
v.t- ■■■»
.jo ■
j'*
29
3c Steady-State Period
The water-ice and CCl^ - dry ice cold traps were in¬
stalled during the start-up period, and if a liquid feed rate
larger than 50 cc/hr. was used, the cold water condenser was
put into operation to remove some of the heat load from the
cold traps. For most runs the duration of the steady state
period was 1 hr., but at higher flow rates only 1/2 hr. periods
or less were used. At the beginning and end of the steady
state period, the level in the calibrated feed reservoir was
read so that the volume of feed used could be determined.
The temperatures in the reactor as well as the con¬
troller bath temperature were recorded throughout the run.
The temperature at the exit line was checked every ten minutes
with the attached potentiometer and the input to the line
heater adjusted if required.
In Appendix II a complete temperature history for two
experimental runs is shown. The first set is for an experi¬
ment with a n-propanol feed rate of 75 gm/hr. and a reactor
temperature of 461 e C. Under these conditions, a 5°C tempera¬
ture difference appears to exist between wall and center-
line of the reactor. The second set shows the worst case, at
a very high feed rate, may have a temperature difference as
high as 15°C. The large temperature difference results from
the large amount of heat consumed in the dehydrogenation of n-
propanol to propionaldehyde. Because the heat transfer through
the catalyst bed in the reactor is not sufficient, this results
* -
: .. : 4 ■
f ■ , t>m ..-. \ :.v /*!
i '"' ’ \ a- 0€ atnj :r .• ,, r:
" - ' 3 ■-
' ' - 1
1 ' ' - '■ '-7 . .... ;
U ** .... ...
qA til
.
■
, ■
“ - : it vo- . , r
'
:
- t. b‘ d Jc , ■ ; ;. J
30
in radial as well as axial temperature gradients. The axial
temperature gradients never exceeded 1°C per inch and were
usually less than 0.25°C per inch.
The pressure in the reactor was read on the attached
mercury manometer and regulated within ±1 psi by adjusting
the throttling valve.
The total liquid product formed in this period was col¬
lected but only 1.5 liters of gas products were collected for
analysis, the remainder of the gas being metered by the wet
test meter. The gas sample was collected by withdrawing brine
through the bottom valve and the gas sampler at about 30 cc/min.
The gas sampling was started 10 minutes after the beginning of
the steady state period to allow the air initially in the cold
traps to be purged from the system so that a minimum amount of
air entered the gas sampler. The pressure and the temperature
of the gas were read with a manometer and thermometer located
on the wet test meter.
4. Shut Down
The product collecting system was isolated from the
reactor, and the liquid feed and heater shut off. As soon as
the liquid feed was shut off, nitrogen was passed through the
reactor. This purged the liquid feed remaining in the lines,
and also maintained the pressure within the reactor during
the cooling of the bath. The nitrogen flow was shut off
after one hour but the pressure in the reactor was maintained
$S.x ' v ■ Bg •: ■■ 2 £ ■ . < (
■' • rrc*£
'
tnotttv.
a-""' Ov ithlA r 1 : tAAviS ' £k#d$ : '&t$T
■ ' . ' ■■:'■■
'•' ■■■ ■
Uj.r: . ?.<a :3
' ' - . ' ■ : - 1 T;' )• *«>■;• i *3 f ■ " & ; ' ■ : ' ; &g fc.fi \ v.g
S:?.j
•, :x?;i 'c fa m&ir*). :■.■■■;■■< sq ' &d ; 5i 'aq&'ij
o'tf, .
■ l
■' ;■ • ' '■ ■■ .. ;■ . ' ;1 /g
.
<■.■ ; ■' ■ ■ ■■■ ■”■ ■■ -'■> pfvi.:.; - ;
■' ■ ■ : • ■■ t ■- r. L ' ,v, ■ ■ ; rv:
= - v ;I.~ I'"-- ' - ' f -'. -j?.? 3 ' + > ' i, ; .5 r.J*^.n a&Ls hri&
' ; r ' " ■ . . :"i - >r T .. ^ *.■ .p ■ ;.r pS
. • •, ■ : ;j. ■; 5 ,
31
until the cerrobase in the bath had solidified. This usually
required about 7 hours after the calrod heater had been turned
off.
While the equipment was cooling, the cold traps were
detached and weighed.
C. Raw Materials
The nitrogen used during the operation of the equipment
was industrial grade (95% pure) nitrogen supplied in cylinders
by the Liquid Air Company.
Only one type of alumina was used in this study and it
was supplied by the Norton Company of Canada Limited. Table 1
gives the physical and chemical properties of this catalyst
as specified by the manufacturer. This catalyst is an impure
alumina containing 21.2% Si0 2 and 1.8% of other metallic
oxides, and for this reason will in the future, be referred to
as Alundum, (the Norton Company Trade-mark).
Sodium hydroxide flakes, purchased from Fisher Scienti¬
fic Company (Catalogue No. 5-312), dissolved in distilled water
were used to treat the Alundum catalyst.
In all but two of the experimental runs, n-propanol,
bought from Fisher Scientific Company (Catalogue No. A-414),
was used as the feed material. The other two runs were car¬
ried out using propionaldehyde and 3-pentanol as feed. Both
of these reagent grade chemicals were supplied by Eastman
Organic Chemicals (Catalogue Nos. 653 and 1355, respectively).
'•••'= : -'C
32
Table 1
Properties of Catalyst
Manufacturer: Norton Company of Canada,Limited
Manufacturer's Designation: LA-617
Properties as specified by manufacturer:
1. Chemical (Typical) Analysis:
A1 2°3
ss
77.0%
SiC>2
=
21.2%
*Fe 2 ° 3
=
0.2%
Ti0 2
=
0.5%
Na 2 0
=
0.5%
to
O
=
0.2%
MgO
=
0.2%
CaO
=
0.2%
Zr0 2
=
-
*Note: All elements are reported as metal oxides.
The actual presence of the metals is in the
form of complex silicates and/or aluminates.
2. Physical Properties:
Porosity
Water Absorption
Bulk Density
App. Sp. Gravity
Packing Density
Surface Area
Max. Service Temperature
X Ray Analysis
60 - 65%
51 - 56%
1.1 - 1.2 gm/cc
3.0 - 3.2 gm/cc
46 lbs/ft 3
60 - 70 sq. meters/gm
700 °C
predominantly Gamma Alumina
some Alpha Alumina and Quartz
■
j ■ •• ■
: , : hkv it,'-:-
ais'-s - r :
33
The purity of the three feed materials was checked by gas
chromatography and only traces of impurities were detectable
in any of these feeds.
D. Analysis of Product
1. Liquid Products
The components of the liquid product were separated
by gas-liquid chromatography using a Burrell Model KD Kromo-
tog. The following is a description of the column used and
the operating conditions under which the liquid analysis was
carried out.
Column:
Column
Temperature:
Sample Size:
1/4 inch diameter# 25 foot# spiral
copper tubing packed with 30% Ucon on
Celite. The Celite was obtained from
the Burrell Company# the Ucon# desig¬
nated as Ucon Lubricant LB-1800X was
supplied by the Union Carbide Chemical
Company.
100°C for five minutes after sample
injection followed by temperature
programming at 15°C/min. up to 150°C.
4.0 microliters
Carrier Gas: Helium
Gas Flow Rate: Carrier side - 60 cc/min
Reference side - 30 cc/min
The products separated and identified under the above
conditions were: 2-pentene# propionaldehyde, methyl ethyl
.
<■■■' .« ' X •: h XX.: ■ ; ' ' : X. ■ Xx\,'^ 'XOXX ■ XXX
■ ; ; ' ' , X...X, ;• XvX X : X
BXsaDO^X ^.. y x i J , XX
:.. ; •: ,ov ■ \ , >.* ■, x XX' XX :' V • ; x, xxX
'
-
.: ' x ■ x: x.., .-'o • ■ X
■ : .
X-
: X?, tr. X:-|n: ,a
XJX2 =CX' :■ X
■ ' . .. •: X: X ’TT ' ' ■
34
ketone, n-propanol, diethyl ketone, ethyl isopropyl ketone,
3-pentanol, n-propyl propionate, 2-methyl-2-pentenal, and 3-
heptanone. All of the above compounds, except 2-pentene and
3-heptanone,were identified by Vasudeva(32) using chromato¬
graphic retention times, infrared spectroscopy and/or nuclear
magnetic resonance. The 2-pentene was identified using reten¬
tion times only. Vasudeva(32) reported the presence of 3-
octanone in the liquid product on the basis of NMR results.
Upon obtaining a pure sample of 3-octanone from K & K Labora¬
tories, it was found that neither the NMR spectrum nor the
chromatographic retention time of the 3-octanone agreed with
that obtained from the product sample. Good agreement was
found with 3-heptanone.
A table of the retention times of the known compounds
along with a typical chromatogram from a liquid analysis may
be found in Appendix III. This chromatogram shows that be¬
sides the identified compounds, many small peaks, at times
amounting to 10% of the total area, are separated by the column.
The composition of the liquid product was found by
comparing the chromatogram obtained from the product to a
chromatogram obtained from a synthetic mixture of known compo¬
sition, approximating that of the liquid product being analysed.
The approximate composition of the product was estimated by
doing a preliminary analysis and estimating the weight percent¬
age of each component from its area on the chromatogram.
'
il ; , 1 »■»*{&' •: (. ■? cauift Oi W-Ji
■ 8 - !
1 ' 1 lo ' rtO J'ttfeo cr ,5« : £.' J ; . „ . oc ,
■’ ■ &VX>i: - 1 -^O*- ' •:'»/ ■ e., V, ,... si ■ , ; : | f ,r
35
The compounds used to prepare the synthetic mixture
were propionaldehyde, methyl ethyl ketone, n-propanol, diethyl
ketone, 3-pentanol and n-propyl propionate. These six com¬
pounds always amounted to more than 70% by weight of the liquid
product. A calculated amount of iso-propanol was added to the
synthetic mixture so that the weight percentages of the other
six components was equal to that estimated from the preliminary
product analysis. Iso-propanol was used since it did not inter¬
fere with the separation by the chromatograph of any of the
other substances present.
The composition of the unknown sample was then calcu¬
lated by using the following relationship:
weight % of A in unknown = (weight % of A in synthetic
mixture)
Peak area of A in unknown
x (------)
Peak area of A in synthetic mixture
The area of each peak was taken to be the height of
the peak multiplied by its width at half the height. All com¬
positions were calculated as weight per cent, and are reported
in Appendix IV.
For the components that were not present in the syn¬
thetic mixture, but which were present in the product sample,
such as 2-pentene, ethyl isopropyl ketone, 3-methyl-2-pentenal,
and 3-heptanone, relative thermal response was used to cal¬
culate the amount of these compounds present in the product.
■ : ■ v ; .??- : -:/i v D'V':;
.
■ :. : 1 ■ /. . i" ■■■-: ;C
■ ; .. - -• : • ■ •» '
ili
. .
do ; v,’ ®uii I ’
1
36
Messner et al and Rosie et al(l6,24) report the relative ther¬
mal response of a large number of organic compounds. Messner
et al(16) also present a method of calculating thermal res¬
ponse factors for compounds if thermal response data is avail¬
able for two other compounds in the same homologous series.
The celite-ucon column was not suitable for water
analysis, since the water peak was not reproduceable due to
excessive tailing. For this reason Karl Fischer titrations(29)
were performed on the liquid samples to determine their water
content. The titration was done by carrying it to a color end¬
point. This is not a very reliable procedure and easily sub¬
ject to error.
2. Gas Products
Since no single column could be found to separate
satisfactorily, all of the components in the gaseous product,
three separate columns were used to analyse the gas. The fol¬
lowing is a description of the columns, the operating condi¬
tions used, and the products analysed with each column.
a) Column: 30% Udon on celite (same as for
liquid product)
Column
Temperature: Room Temperature (24° - 27°C)
Carrier Gas: Helium
Gas Flow Rate: Carrier side - 20 cc/min
Reference side - 30 cc/min
Components
Separated:
Carbon dioxide, propane and propylene
■ - 1 J '■ •- -A ’ ' ' 'J , :!; ■ f.;7<' ? *?.i
£ & a
-v. :r, aq ■ >vl v.
:'t ■. ■.,.. 7 r .1 ,n
' ■ ■ ■■ i'"/ • O : •
... - - f •' v ■ 7~'
■, .,\ r : 'iv.i--.a.'
-
i ,
'■ ' ' ■ :. : ' f ' ; il ■■■■.: . -
r n ■ •• A« ■ V j ’ ■ s-■'£> no *. -ni:/.> -■.:■ \ ■; ■ -■,$jj
- f : " ■ ■ ■•-o '. : .. '■ ■ ... ■ T
• ' ' ' ■ • ■; ; v ■
■
. , IV \
j-; j ; 1-jC ,
37
b) Column:
3/16 diameter 8 foot spiral stainless
steel tubing packed with high activity
charcoal. The charcoal was supplied
by the Burrell Corporation (Catalogue
No. 341-10)
Column
Temperature:
Room temperature (24° - 27°C)
Carrier Gas:
Helium
Gas Flow Rate: Carrier side - 40 cc/min
Reference side - 30 cc/min
Components
Separated:
Hydrogen, oxygen, nitrogen, carbon
monoxide and methane
c) Column:
1/4 inch diameter 25 foot spiral copper
tubing packed with silica gel contain¬
ing 3% di-n-decyl phthalate. Silica
gel was supplied by Burrell Corporation,
(Catalogue No. 341-144). The di-n-decyl
phthalate was purchased from Eastman
Organic Chemicals.
Column
Temperature:
Initial column temperature 100°C.
Temperature programmed 8 minutes after
sample injection at 15°C/min up to 150°C
Carrier Gas:
Helium
Gas Flow Rate:
Carrier side - 24 cc/min
Reference side - 30 cc/min
Components
Separated:
Ethane, ethylene, propane, propylene,
butane, and butenes.
The products were identified by their respective re¬
tention times obtained by passing pure samples of the gas in¬
volved through the appropriate column. For quantitative
' B soVjS .£» $1 .
'■•"■W- s.irrt-'iMZ -.mj&'d
' " ■- ^ f ' •;.:; ... 7 7 ,.';.
■v- - • • , ::.,
...
IV: Jtp0
■ - ' 77
.TV,.:-"
;l -. f 7
1 '■ f 1 >•, L i M
' " 3 ' * v*L*nu vc , s ,, u . }
38
analysis, calibration charts of peak area versus mole per cent
were prepared for carbon dioxide, propylene, nitrogen and car¬
bon monoxide. The graphs were obtained by analysing gas samples
of known composition and plotting the peak area obtained versus
composition. For hydrogen, a modified peak height criterion
was used since an area under the hydrogen peak could not be
found because at high concentration of hydrogen two peaks re¬
sult (see hydrogen peak in Figure 3 in Appendix III). The
amount of the other gases such as methane, ethane, ethylene,
and propane, was found by using relative thermal response
data(16,24) .
To check the gas analysis, several gas samples were
analysed on the chromatograph and by an orsat apparatus equipped
to determine carbon dioxide, carbon monoxide, hydrogen, oxygen
and unsaturated hydrocarbon content. These comparisons are
given in Appendix III.
3. Accuracy of Analysis
a) Liquid Products
For those liquid compounds that were included in the
synthetic liquid mixture used as a reference for analysis, the
error in the reported values of compositions is not more than
t2%. For the other liquid components where relative response
was used, the error is probably not much larger, but no
means of checking it were available.
' « ■ - - m r , ■ ■
■
39
b) Gas Products
The values reported in Appendix III comparing the gas
analysis by gas chromatography to orsat analysis are not in
good agreement. It is believed that the gas chromatographic
analyses are more accurate and that the gas compositions do
not vary by more than i5% of the reported value.
With the available equipment it was difficult to pre¬
pare seven component gas mixtures with accurately known com¬
positions. For this reason no synthetic mixtures were prepared
to check the gas analyses.
o:> 2*0 riay L-iisr,
' • ' . - r - - -■ .. : ■ ■ «*v .
■
'
. ■ '
: • r ■ '1 - ■
40
IV. EXPERIMENTAL PROGRAM AND RESULTS
It was decided to prepare a base-exchanged catalyst
using the Alundum obtained from the Norton Company, as described
by Vasudeva(32), and if the 42.2% yield of diethyl ketone, he
reported could be reproduced, to investigate further, the acti¬
vity of this catalyst. This investigation would involve the
following objectives:
1. to investigate the effect of stainless steel upon
the decomposition of n-propanol;
2. to determine the activity of the catalyst;
3. to find the effect of temperature and space
velocity on the types of reactions occurring,
and their rates;
4. to find evidence in support of a postulated
reaction scheme; and,
5. to measure the life of the catalyst.
The approach taken, the experiments done and the re¬
sults obtained to accomplish the above objectives are out¬
lined in detail in the following section. The complete set
of operating conditions and product distributions obtained in
performing these tests are presented in Appendix IV.
A set of definitions of terms used in the reporting of
results follows.
-
' *k E»« -
.
41
Definition of Terms
1. Reaction Temperature
The reaction temperatures reported in cases where the
glass reactor was used, was taken to be the arithmetic average
of the two thermocouple temperatures, one located at the top of
the catalyst bed, and the other at the bottom.
For the stainless steel reactor, the reaction tempera¬
ture was computed as the arithmetic average of the temperatures
recorded by the center-line thermocouples which were located
in the catalyst bed. For all the runs except Runs 16 and 17,
a large catalyst charge was used and thermocouples TCI to TC4
(see Figure 3), were located inside the catalyst bed. For
Runs 16 and 17, when only a 15.8 gm catalyst charge was used,
only thermocouples TCI and TC2 were located inside the bed.
2. Space Velocity
The term space velocity as used in this work is de¬
fined as
gm-moles of n-propanol feed
Space velocity = -
(hr) - (gm of catalyst)
The term moles will be used in the remainder of the work and
will always refer to gram-moles.
3. Diethyl Ketone Yield
The maximum yield of diethyl ketone is one mole per
two moles of n-propanol reaction. It will be shown in a
" . n. J.J: v,v?;s
-
'HU ■ "j i' $r' \
* " ■ -
' *' ■■■■■■ «.
' :
■
■
'
■ .
-V: sfjr-.+fO' f.,
' ' ■ : .u •, : ft,:-
■ ' ' ' 2
42
later section that there are reaction sequences which require
three moles of n-propanol to yield one mole of diethyl ketone.
But to avoid ambiguity, the diethyl ketone yield is defined
as:
2(moles of DEK in product per mole of n-PrOH fed)
% DEK yield = - x 100
(n-PrOH conversion)
where
DEK = diethyl ketone
n-PrOH = n-propanol
moles of n-PrOH in product
n-PrOH conversion = 1 - (-—---)
moles of n-PrOH fed
4. Material Balance
The reported material balance values are a measure of
the efficiency with which the products could be collected.
Material
Balance
(weight of liquid product)+(weight of gaseous product)
(weight of liquid feed)
For the calculation of moles of product per mole of n-propanol
feed, the weight of the liquid product was adjusted so that a
100% material balance was obtained. Here the assumption was
made that the collection of the gaseous products was much
more efficient than that for the liquid products.
£f> •
T
. j .
y. ■ umh
■ ■ ■ . <’
■ • ■ ■ ' i no,, v
- ;
■
.
• ■■ ■ • >». . • . j
' * ChO ■ ’ ■
. ' l . '
■
■ '■bst-o'.r hoi'.toqo-x g ;fT
■‘•■'Xy- ■ r: } ‘ h„; ■ -vv J- ■ , .. ,
' • ' ;©
. < ■ •- '. : .... i •!.
■ h ;o, Ch
' ' ' " ' ' ; o . : ,, ..
■
' “ ■ “■ -
■ - |>h; Slji , . fK:ji
'
43
5. Material Accountability
Material accountability was used as a measure of the
completeness of analysis of the products. It is a carbon,
hydrogen and oxygen balance.
% Carbon accountability *
gm-atoms of carbon in analysed products per mole of n-propanol feed
gm-atoms of carbon in 1 mole of n-propanol feed
x 100
similarly for hydrogen and oxygen.
A. Initial Studies
1. Preparation of Catalyst
The catalyst used was an Alundum catalyst carrier sup¬
plied by the Norton Company of Canada, the chemical analysis
of which was cited earlier.
The Alundum catalyst particles used were of irregular
shape, having a size range of -10 to +20 mesh. These particles
were completely immersed in a 2.0 N sodium hydroxide solution
for 6 hours at room temperature. After 6 hours, the excess
sodium hydroxide solution was removed by filtration, and the
wet catalyst particles dried in air at 95 to 105°C for 5 hours.
Two 250-gm batches of catalyst, C-l and C-2, were pre¬
pared in this way, so that the reproduceability of this method
of preparation could be checked.
.
■
■ '■ J ' -
: ' '■ ■ • : - : ,,
u :;J ■ i/o
44
An analysis for the sodium content of the NaOH treated
Alundum catalyst was carried out by the Research Department
(analytical section) of Chemcell (1963) Limited, Edmonton,
Alberta. The sodium content of the treated catalyst was found
to be 3.3% by weight sodium, using flame photometry as the
method of analysis.
2. Activity of the Catalyst
In order to determine the activity of the above described
catalyst, as well as the reproduceability of the catalyst pre¬
paration, three experimental runs were carried out. Runs 1
and 2 each used the glass reactor and similar operating con¬
ditions but with catalyst C-l and C-2, respectively. The
glass reactor was used so that the results could be compared
to those reported by Vasudeva(32). Run 3 was done on the
stainless steel reactor using catalyst C-2 and a temperature
and space velocity similar to those of Runs 1 and 2. The re¬
sults obtained are reported in Table 2.
For all the runs using a catalyst, other than Run 1,
the catalyst C-2 was used.
3. Stainless Steel as a Catalyst
The results of Runs 2 and 3 indicate that the decompo¬
sition of n-propanol is affected by the presence of stainless
steel. To investigate the magnitude of this effect, a series
of 3 runs, Runs 4, 5 and 6 was made. Run 4 was carried out
on the stainless steel reactor without a catalyst charge.
nw ■. ■■■■ :!£■ ■ . . . ■ . f..-
' ' ■ : •* :■ : . :
• . l
■
45
Table 2
Performance of Base-Exchanged Catalyst
Vasudeva
Run No.
1
2
3
Run 10
Catalyst Used
C-l
C-2
C-2
-
Reactor
Glass
Glass
Stainless
Steel
Glass
Reaction
Temperature, °C
465
470
464
461
2
Space Velocity x 10 *
0.768
.697
.615
.667
n-Propanol
Conversion, %
95.5
94.8
93.6
95.8
Diethyl Ketone
Yield, %
40.9
41.0
45.8
42.2
Material
Balance, %
93
94
95
84.6
Material Accountability^
Carbon
74.3
72.3
82.3
-
Hydrogen
78.6
78.4
85.3
-
Oxygen
99.2
99.1
99.5
-
* The units of space velocity as used in
moles of feed/(hr)(gm of catalyst)
all tables are
-
s dr . :■
.
'
c c
'
•
.. V ;
,
■ - ;,;u ' :&:■ ■ i £■ ■ •:!
•• ■ " I •
' ■/ ■ ■-’ '!/. CJi
46
To determine whether the decomposition of n-propanol observed
in this run was due to the stainless steel or only due to
thermal decomposition, the magnitude of the thermal decomposi¬
tion of n-propanol was obtained by carrying out Run 5 on the
empty glass reactor. In Run 6, propionaldehyde was used as a
feed to the empty glass reactor and data as to the thermal
stability of propionaldehyde was obtained from this run. Table
3 summarized the results for these three runs.
B. Influence of Temperature and Space Velocity on Catalytic
Activity
1. Effect of Temperature
A set of 6 runs. Runs 7 to 12, was designed to investi¬
gate the effect of temperature on the dehydrogenation of n-
propanol over a base-exchanged alundum catalyst. The tempera¬
ture range considered was 300 to 500°C and runs were carried
out at 40°C increments with the space velocity maintained at
approximately 1.04 x 10 moles of n-propanol/hr-gm of catalyst.
Table 4 gives the operating condition and summarizes the re¬
sults for these runs, while Table 5 and Figure 5, 6 and 7
show the liquid and gaseous product distributions and material
accountability.
2. Effect of Space Velocity
The effect of space velocity on the dehydrogenation
of n-propanol was determined by doing 6 runs, Run 3 and Runs
13 to 17, during which the temperature was held constant at
V
■ anA: ‘ 'svtoBc-x amx% &qm* <•■ j- a fc.-f>2
i
[
- .
• ■■■■■ OJ 6,0 e ■ ■ r TO » u <* V sma , aruj'x ; j c JM A
'
* ^ * •*' 8BW **»«•*■* ^ :» doiw p.,i u o:> «
Effect of Stainless Steel and Glass on
47
0
0
G
G
0
o
-P
-p
0 0
0 0
0
G
T) «
TJ «
T3
O
>i
•H
•H
X rH
X rH
X
■p
to
0 >1
0 >i
O
o
-P
rf x
T5 X
G
0
o
rH -P 0
<H -P
0
Q)
G
(DW C G
0 W G
s
PS T3
G 0 0
G 0
0
O rH tP ' — 1
OH tn
0 G
G
P
•H >i 0
•H >iO
G 0
•H
a
ax p a
ax p
0 X
fd
O -P T3 O.
o -P V
X p
s
p 0 >* p
p 0 >i
-P 0
a S k a
a s «
W U
rH
o
G
1X3
a
o
u
a
I
G
o
G
O
•H
-P
‘H
W
O
a
g
o
o
0
Q
<D
X
-P
G
O
•H
(0
U
a)
o
u
a)
u
G
-p
0
p
0) u
a«
g
<u
Eh
* o
h a)
to
-P
1X3
« U
X
\
a) g
0) CP
En
03
0
0
Eh
Sh
O
-P TJ
a o
rd to
O p
§
PS
00
O
•
•
•
VO
CM
CM
o
O
o
IQ
<7\
rr
•*?
cn
<T>
CM
CM
rH
CM
in
CO
rH
•
•
•
in
00
00
rH
0)
x
Q)
«H
rH
T)
o
o
■—1
G
G
0
0
0
G
a
a
o
o
o
*r4
p
u
a
04
1
04
o
1
u
G
G
04
to
to
0
G *H
to
to
•H 0
to
to
0 0
0
0
-P -P
rH
rH
to to
u
u
H 1
in
ID
l
t-
*
Residence time, based on empty reactor volume
. .• :
.
s.
1 ■ te»
■V ■ v,: J
V,. U
- '■
•
' i -
• . •
i '• *„»
' -
CO
r
t-,
■
i-'l
i£i
iv
: ! '
;
.
o, :
a
c : ki
V - .
■
Effect of Temperature on Catalytic Activity
48
<d a)
•h o
p G
0) id
-P rH 0\0
(0 id
£ pq
CTt 00
r~ vo vo
CTi
CTl CTi
<T\ CT> O
O G
G O
id *H
a w
0
p
■'3*
CTi
fH
CM
H
p
Q>
•
•
«
•
•
•
>
O
rH
o\
in
r*
1
G
H
m
o
<1)
p
G
-P
id
P u
rH
< 7 >
o
G\
CT\
'JO
Q) o
O
CO
CO
1—1
in
o>
a
CO
CO
CO
* 3 *
•^r
e
(D
>i
•P
•rH
O
O
i—KN
a) o
t> rH
x
0)
u
id
a
tn
CM
O
CM
co
in
in
CO
*3*
CO
•'3'
CO
CM
o
O
O
o
o
o
•
•
•
•
•
•
rH
rH
rH
H
rH
H
Q)
p
G id
00
00
CM
•<3*
rr
•<3*
W -H
W W
rH
1—1
CM
CM
CM
CM
<d a
p
pj
co
CO
CO
co
CO
CO
49
Table 5
Effect of Temperature on Product Distribution
(moles of product per 100 moles of n-propanol feed)
Run No.
Product
7
8
9
10
11
12
2-pentene
-
-
-
0.24
0.61
3.80
propionaldehyde
0.14
0.91
2.62
4.90
4.30
0.80
methyl ethyl ketone
0.034
0.16
0.26
0.58
1.12
1.94
n-propanol
99.64
98.11
92.90
80.80
44.58
2.87
diethyl ketone
-
-
0.26
1.84
10.50
17.95
ethyl isopropyl ketone
-
-
-
0.13
1.16
1.85
3-pentanol
-
-
-
0.11
0.50
0.45
n-propyl propionate
0.058
0.153
0.86
1.14
0.43
0.14
2-methyl-2-pentenal
-
-
Trace
0.10
0.13
0.38
3-heptanone
-
-
-
0.14
0.44
1.61
water
Trace
0.2
0.5
2.0
3.0
3.1
hydrogen
0.35
1.90
7.65
20.72
63.88
124.0
carbon monoxide
-
-
0.13
1.27
8.06
21.4
carbon dioxide
-
-
0.35
1.54
10.15
17.6
propylene
-
-
0.10
0.62
2.25
5.29
propane
-
-
0.13
0.25
0.41
1.54
ethylene
-
-
-
Trace
Trace
Trace
ethane
-
-
0.03
0.15
0.52
3.40
methane
-
-
0.03
0.10
6.61
1.70
MATERIAL ACCOUNTABILITY
carbon
99.9
99.5
98.4
94.9
83.2
73.6
hydrogen
100.0
99.7
99.0
96.5
86.3
80.0
oxygen
99.9
99.7
99.1
97.1
94.8
87.5
'■ -'JV •/
- ■ : ,:.P. Of; "
- .y _' : V\ ^ to ' ..,j ■
0 , :
:
■ . . fj
?. :■ s'tT
7-ai
ui ; PO"i v
■
OOB.u::
r ,• *,
» r. i
i
. », y.; ■
■<
MOLES OF PRODUCT / MOLE OF n-PROPANOL FEED PERCENT
50
AND DIETHYL KETONE YIELD.
FIGURE 6
EFFECT OF TEMPERATURE ON LIQUID PRODUCT DISTRIBUTION,
(•pact velocity = 0.01035 moles of n-propanof / hr. - gm. of catalyst)
*
.
■- U vr " y . >S . 'J
MOLES OF PRODUCT / MOLE OF n- PROPANOL FEED MOLES OF PftOOUCT/ MOLE OF a-PROPANOL FEED
51
FIGURE 7 A
FIGURE 7 s EFFECT OF TEMPERATURE ON GASEOUS PRODUCT DISTRIBUTION
(spoct velocity *0.01099 molo* ot n-proponol /hr.-gm. of catalyst)
{•«■■ - r .r
-
'
52
463°C and the space velocity varied from 0.615 x 10 2 to
_2
19.16 x 10 moles of n-propanol/hr-gm of catalyst. The rea¬
son for not investigating a wider range of space velocities
was due to limitations in the construction of the stainless
steel reactor. With the feed system employed, it was not pos¬
sible to control flow rates accurately below 25 gm/hr. The
product collection system placed an upper limit on the flow
rates since it could only condense the vapors adequately if
the n-propanol feed did not exceed 190 gm/hr.
A catalyst charge of 71.4 gm was used for Runs 3, 13,
14 and 15, but to obtain higher space velocities of Runs 16
and 17, the catalyst charge was reduced to 15.8 gms. New C-2
catalyst was used for these last two runs. To reduce the
amount of catalyst further may result in the decomposition of
n-propanol due to the stainless steel becoming more important
than that due to the catalyst.
A summary of operating conditions and results for
these runs is given in Table 6, product distributions for
these runs are presented in Table 7 and Figures 8, 9 and 10.
Upon examining these tables and figures, it can be
noted that the results for Run 3 seem to be in error, the val¬
ues for the composition of the liquid components all seem too
large. This is also evident in the values for the material
accountability. The probable reason for this is that the
liquid sample injected in the gas chromatograph was too large.
" ’ " ■' ' ; ' ' ■'
Effect of Space Velocity on Catalytic Activity
53
Cd Q)
-H O
H C
0) id o\°
-P i—(
cd cd
S CQ
in vo vo
CV CT\ CT\
C\ O
CT> O
CD
G
O
-P
CD
T3
00
o
CM
00
C5>
-O'
1—1 1—1
•
•
•
•
•
•
>1 CD
in
00
VO
rH
00
ro
x -h
-3*
ro
ro
ro
rH
•—\
-P >H
CD
Q
rd "H
a cn
VO
VO
o
00
VO
o
o u
M CD
ro
00
CM
CM
rH
vo
>
cn
vo
in
CM
1—1
1 G
G O
U
>i
-P
•H
U
OCN
i-H O
CD <H
in
co
in
00
>
iH
VO
CM
rH
O
r'-
X
VO
rH
00
00
VO
rH
CD
O
o
*H
rH
CM
<j\
fd
»H
a
w
CD
U
P
-P
cd
u
CD U
CD
E-i
ro
i—l
I—1
ro
vo
vo
vo
vo
vo
vo
*3*
-S’
CD
U
U
CU
P
cd
00
CM
rH
vo
vo
VO
(0
•H
•
•
•
•
•
•
in
in
o
CM
CN
CM
CM
CM
CD
a
ro
CO
ro
ro
CO
CO
O
s
n
ro
co
in vo r"
<—C *—I rH
/■is
•
t<i ., -
r, cn
■ ■
«■ (-■
v -
'
54
Table 7
Effect of Space Velocity on Product Distribution
(moles of product per 100 moles of n-propanol feed)
Run No.
Product
3
13
14
15
16
17
2-pentene
1.50
0.46
0.30
0.30
Trace
-
propionaldehyde
2.03
4.58
5.18
6.28
8.11
8.25
methyl ethyl ketone
2.49
1.53
1.40
0.96
0.59
0.44
n-propanol
6.44
31.40
48.00
57.20
78.40
84.00
diethyl ketone
21.50
13.00
9.42
6.81
2.04
1.07
ethyl isopropyl ketone
1.79
0.62
0.87
0.66
0.11
0.05
3-pentanol
1.07
0.60
0.45
0.24
0.06
-
n-propyl propionate
0.28
0.45
0.72
0.87
1.32
1.24
2-methyl-2-pentenal
Trace
0.20
0.23
0.39
0.06
Trace
3-heptanone
2.30
1.25
1.21
0.90
0.13
0.05
water
2.5
2.6
2.6
2.3
1.1
1.1
hydrogen
119.0
85.0
72.6
58.61
27.60
19.06
carbon monoxide
20.3
10.20
6.95
4.08
1.64
0.75
carbon dioxide
19.1
14.2
7.70
5.02
2.01
1.35
propylene
5.10
3.29
1.32
1.16
0.68
0.48
propane
0.61
0.43
0.25
0.18
0.13
0.09
ethylene
Trace
Trace
Trace
Trace
Trace
Trace
ethane
1.22
0.70
0.25
0.19
0.30
0.15
methane
1.80
0.98
0.41
0.30
0.37
0.32
MATERIAL ACCOUNTABILITY
carbon
,%
82.3
79.1
84.7
86.8
96.5
98.8
hydrogen
85.3
83.8
91.2
92.7
98.4
99.8
oxygen
99.2
95.1
93.0
91.2
98.9
100.0
-
G-
'■ ' ■ » ■' ' • ' r-.r
-
MOLES OF PRODUCT/MOLE OF n-PROPANOL FEED PERCENT
55
FIGURE 8 = EFFECT OF SPACE VELOCITY ON n-PROPANOL CONVERSION
AND DIETHYL KETONE YIELD
FIGURE 9 : EFFECT OF SPACE VELOCITY ON LIQUID PRODUCT DISTRIBUTION
; . u# •■■■■ ;
MOLES OF PRODUCT/MOLE OF n-PROPANOL FEED MOLES OF PRODUCT/ MOLE OF n-PROPANOL FEED
- 56 -
FIGURE 10 B
FIGURE 10 : EFFECT OF SPACE VELOCITY ON GASEOUS PRODUCT DISTRIBUTION
(temperature = 463°C.)
'
■
■
57
C. Special Runs
One method used for studying reaction schemes, is to
use intermediate products occurring in the postulated sequence
as feed and then to observe whether the products obtained are
the same as the ones predicted.
Vasudeva(32) reports the product distribution for pro-
pionaldehyde, n-propyl propionate and diethyl ketone as feed
over a base-exchanged chromia on alundum. This catalyst has
a very similar activity and selectivity as the base-exchanged
alundum by itself in the dehydrogenation of n-propanol, this
can be seen by comparing the product analysis obtained by
Vasudeva with those in this work for test with similar operat¬
ing conditions. This comparison is presented in Table 8.
Two other compounds, whose reaction products over
this catalyst would be of interest, are the aldol of propion-
aldehyde and 3-pentanol. It was not possible to do a run
with the aldol, since it is not very stable and hence dif¬
ficult to obtain. Run 18 was made using the glass reactor
and 3-pentanol as feed. Table 9 reviews the results obtained
by Vasudeva with the special feeds and also presents the
results of Run 18.
D. Poisoning of the Catalyst
The color of the catalyst changed with use from a near
white after preparation to a dark grey after about 40 hours
' p fl , .
; , : O il i j “
’ • * •• ■
9 - •• Y ,., J , S3 sri3 sa ^
58
Table 8
Comparison of Catalytic Activity of Base-Exchanged
Alundum to Base-Exchanged Chromia on Alundum
Vasudeva
Run
2
Run 17
Temperature °C
470
467
Space Velocity
moles/hr-gm of catalyst
0.697
0.617
Product Distribution (moles per
100 moles
of n-propanol fe<
Propionaldehyde
1.21
1.01
Methyl Ethyl Ketone
2.38
2.00
n-Propanol
5.24
0.77
Diethyl Ketone
19.55
23.60
Ethyl Isopropyl Ketone
2.08
2.00
n-Propyl Propionate
0.17
0.13
3-Pentanol
0.43
0.46
Hydrogen
126
147.0
Carbon Monoxide
27.7
24.5
Carbon Dioxide
18.7
23.1
Propylene
5.2
4.7
J
;,c
, r ■
59
Table 9
Operating
Conditions and
Product Distributions
for Runs with
Special Feeds
Run No.
18
20*
21*
22*
Feed
3-Pentanol
Propionaldehyde
n-Propyl
Propionate
Diethyl
Ketone
Space Velocity**
0.72
0.32
0.40
0.43
Temperature °C
462
457
456
467
Product Distribution (moles per 100 moles of feed)
2- Pentene 18.1
Propionaldehyde 0.0
Methyl Ethyl Ketone 2.92
n-Propanol 0.0
Diethyl Ketone 75.3
Ethyl Isopropyl Ketone 1.34
3- Pentanol 0.80
n-Propyl Propionate 0.0
Water 6.0
Hydrogen 98.2
Carbon Monoxide 2.34
Carbon Dioxide 5.00
Propylene 0.0
not reported
0.99 2.56 0.78
1.62 3.16 4.82
Trace 1.42 0.0
21.50 52.70 78.90
1.06 2.61 2.21
0.29 0.61 0.41
Trace 0.33 0.0
not reported
47.6 47.8 8.5
11.6 45.2 0.96
25.6 18.6 7.08
1.4 12.7
* Run as reported by Vasudeva(32)
** Space velocity expressed in gm of feed per hr-gm of catalyst
0.32
; .
a
■
V'\
:a
t , X
. . i
. - t.
- • ■ : K
60
of use. This color change was due to the deposition of elemental
carbon on the catalyst surface. Runs 19 and 20 were carried
out to determine the effect of this carbon deposition on the
catalyst's activity. In Run 19 an attempt was made to repro¬
duce the operating conditions of Run 3. Run 3 was performed
with new catalyst, while the catalyst had been used for ap¬
proximately 75 hours when Run 19 was made. At the time when
this run was made, it was not known that the results for Run 3
were questionable, and for this reason the comparison of the
results of Run 3 and 19 as found in Table 10 are not as con¬
clusive as would have been the case if the results of Run 3
had been accurate.
To investigate the change in activity with time of
the catalyst further, Run 20 was made using the glass reactor.
The objective of this run was to maintain the operating con¬
ditions constant over a long period of time and take periodic
samples of the product. The duration of this run was 50
hours, but difficulties were encountered in attempting to
maintain constant temperature and n-propanol feed rates. The
only analysis obtained for this run is the n-propanol, diethyl
ketone and propionaldehyde content of the product. The pro¬
duct was sampled every 5 hours. The results are presented
in Table 11 and shown on Figure 11. The large fluctuations
noticed in these values are due mainly to variations in ex¬
ternal conditions rather than a change in catalytic activity.
■
■
£' -- ® .•
■ »i nm 3fy OS , = . .7 1 .. = ■ ..
- . .. . K, :lt
61
Table 10
Life of Catalyst
Run No:
19
Operating Conditions:
Space Velocity
Temperature, °C
Pressure, psia
Length of Catalyst
Use, hr.
0.615 x 10
464
30.8
-2
0.601 x 10
465
31.8
75
-2
Results:
93.6%
74.8%
n-Propanol Conversion
Diethyl Ketone Yield
45.8%
40.2%
62
Table 11
Effect of Time on Catalytic Activity
Run 20
Space Moles per 100 Moles
Time,
hr.
Velocity
x 102 ^
Temp.,
°C
nPrOH
Conversion
of
PrH
nPrOH
nPrOH
feed
DEK
0
0.741
440
38.0
4.72
62.0
11.2
5
0.653
444
38.8
4.20
61.2
11.5
10
0.805
442
64.6
3.01
35.5
14.2
15
0.763
438
43.5
4.48
56.5
14.3
20
0.763
442
34.4
5.00
65,6
10.1
25
0.720
441
41.8
4.33
58.2
12.7
30
0.720
447
65.2
2.96
34.8
13.6
35
0.708
446
79.8
2.43
20.2
19.7
40
0.763
438
32.9
4.88
67.1
9.4
45
0.763
4 43
38.5
4.76
61.5
10.2
50
0.765
443
43.6
4.26
56.4
12.4
where
\
n-PrOH - n-propanol
PrH - propionaldehyde
diethyl ketone
DEK
o -
■■■<■' : . : - V
M ‘
. .
C £
' . ~l
Ui>
Xr ' v f.
. • w.y. ' O ■ ■; . "
v. '. :! , ■
'X
■ ■ ■ *n i'. .. .
1^:0
£, 0
.T y
0£
6 ■
i
PROPION ALDEHYDE
63
bi
Z
6 .
“ 2
-l <
UJ
Q C
I I I
o □ <
<o
o
<£>
—I—
O
m
J
—i—
o
to
—T—
o
(VI
c
C r2
□ C
□ C
□ O
6
□ 6
n
. m
. o
<•
«vj u
2
*■8
m
UJ
oc
3
o
. o
m
0333 lONVdOad-u 30 310W d3d S310W
EFFECT OF TIME ON CATALYST ACTIVITY
::
A
64
V. DISCUSSION AND INTERPRETATION OF RESULTS
The results shown in the preceeding section will now
be analysed and discussed in the following sections:
A. Chemical Reaction Sequence
B. Rates of Reaction
C. Yields
A. Chemical Reaction Sequence
On the basis of compounds encountered in the dehydro¬
genation products of n-propanol, Tollefson(30) and Vasudeva(32)
postulated a reaction scheme which is basic to the analysis
described herein. The following compounds were identified
and analysed for in the reaction product.
1 .
2 -pentene
11 .
water
2 .
propionaldehyde
12 .
hydrogen
3.
methyl ethyl ketone
13.
carbon monoxide
4.
n-propanol
14.
carbon dioxide
5.
diethyl ketone
15.
propylene
6 .
ethyl isopropyl ketone
16.
propane
7.
3-pentanol
17.
ethylene
8 .
n-propyl propionate
18.
ethane
9.
2 -methyl-2-pentenal
19.
methane
10 .
3-heptanone
■ ■ ■ . - . •
5 l.o'i ..•. : c„ : ;
... ■ j :V' .**. *; Xt " ;;.j
. - ... £ x; or '. .": . . ' xv f /x-: x’ ; '
■■'Of:
1 £/ -
65
1. Model of Major Chemical Reaction Sequence
Figure 12 shows a reaction scheme which includes the
majority of the compounds listed above. This reaction se¬
quence, with minor additions is the same as that proposed by
Vasudeva(32).
This reaction system relates those compounds which
were identified in the product. Without doubt, unstable inter¬
mediates such as propionaldol and propionic acid could parti¬
cipate in the reaction sequence as suggested below.
2C 2 H 5 CHO -
-► C 2 H 5 CH(OH)CH(CH 3 )CHO
(25a)
5 CH(OH)CH(CH 3 )CHO -
.- C 2 H 5 CH(OH)C 2 H 5 + CO
(25b)
C 2 H 5 COOC 3 H ? + H 2 0 -
->• C 2 H 5 COOH + C 3 H ? OH
(26a)
2C 2 H 5 COOH -
-► C 2 H 5 COC 2 H 5 + H 2 Q + C0 2
(26b)
Both reactions 26a and 26b are believed to be present
since Tollefson(30) encountered propionic acid as a product.
By examining the product distributions as reported
in Tables 5 and 7, and by noting the sequence in which the
products appear with increasing n-propanol conversion, it is
seen that the reaction route presented in Figure 12 is in
agreement with the data. Only those runs with low n-propanol
conversions, obtained at high space velocities and/or low
temperatures, are helpful in determining the intermediates
in a reaction system.
I ~0k : • ' 6 i rc ■ a . : •. i"i
■ ■' m ■ 20. , f £sr; m . <• '■■■: if :. .
ob . ' • : .'i" ; • /• "■ . - -w;
:a«k; t. - . * >nr* . di r . •
'• v ifo . O'J - • •
-■ t ■ i- \
. q- n p.. ■'■£. . <v
i- •• . jvv vv .• - . - ■;
• £ • ' -...U -: /■ .. : . - :
i . v \ 9v >«qe rIp.:V . £>s*r. , i i . ■ »c
■. '• ■ r £ . ‘ i-.i; , /.< . . ,
•m . . - x .i .
- o u
O
(N
as
10
as
co
U
as
o
co
U
CM
a:
o
as
u
LO
»—«
CM
U
<»
-P
P
s
p
<D
-P
03
W
o
r d
r-H
<3
u
10
as
CM
U
o
u
as
CM
u
as
u
UO
as
CM
u
CM
as
+
o
co
U
as
u
in
as
CM
U
o
CM
as
+
PH
O
r-
as
CO
LO
u
V0
as
co
U
LO
as
CM
O
O -
o
LO
as
CM
U
LO
H-l
»-«
CM
U
O -
U
UO
as
CM
u
CM
as
+
uo
as
CM
u
o
u ~
uo
as
CM
CJ
CM
CO
as
u
as
u
o
u
uo
as
CM
UO
as
CM
u
o
u
co
as
u
o
CM
as
in
as
CM
U
as
u
co
as
u
Figure 12: Reaction Model
67
All of the reactions shown in Figure 12 are assumed
to be irreversible at the conditions encountered in this work.
This assumption is justifiable from the results obtained with
special feeds as reported in Table 9. The results reported
by Vasudeva(32) using either propionaldehyde or n-propyl pro¬
pionate as feed, appear to contradict this assumption since
all the components shown in Figure 12 were present in his pro¬
duct mixtures. It is believed that this is caused by the re¬
generation of n-propanol, and thereby all subsequent reaction
products, from the decomposition of n-propyl propionate either
directly,
2c 2 h 5 cooc 3 h 7 -* c 2 h 5 coc 2 h 5 + c 3 h 6 + co 2 + c 3 h ? oh
or by hydrolysis
2C 2 H 5 COOC 3 H ? + H 2 0 -** C 2 H 5 COC 2 H 5 + C0 2 + 2C 3 H ? OH (12)
2 . Discussion of Reaction Model
The reaction sequence presented in Figure 12 accounts
for the formation of 15 of the 19 identified products. It
is believed that there are numerous other reactions involving
the compounds shown in Figure 12. The following discussion
will show some of the deficiencies of the reaction model.
The absence of methane and ethylene from the reaction
model is not serious since these two compounds occur only in
minor amounts, and probably are the result of thermal decom¬
positions. The propane is not as easily accounted for and
( 8 )
, ■ 5 ; - ' v ..' - ■?! a ' ■ pp. i yr’: ’■ ■■■ ■ -i ;
. \ ‘ h J\;. -tU : d b ; : ■*■ .■::!■ pd O-i
'
' \ ' , i . : : . . . '.■■
■ . ■ :■ ■ / ■ ■ ' vy- y - . ., 1 ■■i-w v •: -.■ i-. . 0
' pypU' \..i PPJp'p :U i'-WOdw IP i 'P PO yr«.X> ' Ysi'rj' V C'
y.yy v. i; s ' ■ - V ■ ib/v- yy ^ i :
y-yy y-. ' ■■ yyy v.yyyyyyy :v"y^y yy\yyy'
■ - ■ . i y ' .. , y. i': • . . ' -
. .
>#crf y yyy
v ' ••• 0 •
'' • i "yb.: tc> tt":P IP t;sf« y ' y i'o!T
■
■d- - -:z ■ ■■• ' • - : •"■■■ -.
'
■ • >y r?)
•'■■■■■■'■ ' " ' . yy. yyyy b:;y yyyyyy p ,< pp -
.. - ■■ ■ '■ ■■ ' - • yy - '■ yy; , : . y ■ ' b,
68
may be a by-product of the dehydrogenation of n-propanol(14 # 28)
The reaction(s) responsible for the formation of 3-
heptanone have not been established, but it is believed to be
caused by further condensation reactions and not by thermal
means. This is indicated by the increasing concentration of
3-heptanone in the product at constant temperature and increas¬
ing time.
The formation of ethyl isopropyl ketone and methyl
ethyl ketone is shown in Figure 12 to occur according to the
reaction
2c 2 h 5 coc 2 h 5 ---ch 3 coc 2 h 5 + c 2 h 5 coch{ch 3 > 2
The above reaction probably accounts for the formation of all
of the ethyl isopropyl ketone found in the product since it
was noted that this compound does not appear in the reaction
product unless diethyl ketone is also present. Reaction 23
accounts only partly for the formation of methyl ethyl ketone
since the product distributions for runs at low temperature
(Runs 7 and 8) indicate that methyl ethyl ketone is formed
before either diethyl ketone or ethyl isopropyl ketone appear.
This suggests that methyl ethyl ketone might be formed by
some reaction involving only n-propanol and propionaldehyde.
That this reaction may be a vapor phase, non-catalytic reaction
is indicated by Runs 4 and 5 where no catalyst was employed
and methyl ethyl ketone was one of the reaction products.
The appropriate reaction has not yet been determined.
(23)
B
'i i ■ 5 ■ ■ 3 IP
. • - ■ . / r ;! U- pzppppxi^ r; '-' r
'■ ;S ■ d-\ ^pp ■; ■
■ .• .. x :-v v ; '' :> '■ • ••• ••
■ ■■' , .:U:.TV■.;■■■ .Xr^ 7. ; v :•» j X's 1 ' ,v v > iu. ■?■■■• Kr.<xP\\'.rxP
■ "v . ■ v ' / * '■ n- - ..av- •?
f c D 0 D r H 9
ft'jtfnaow 1 s
y-!r v : . 7 ,*. - /'-v-i' ■ ■ 1 ; i ■ p :j
. . : .fan a ^xxo<:p;p. • Ppx di ■-%£■*
'■■■ m:P P t> P'"- usix o» c .*s l '{/l:?®ip z-.n*b&*
. rj \ J 3t
■ ••:. ' ■-■• ' ' ': vKoii ■■ h 1 . r -h . dxjBh■■ o ^
I % '
/ . ' : ; '■ - P V PpPP Z ' , ■ P '
--'■P-y-. v^r • *' - ■ pIvJxPPP :P:SPP Z
■ ■ - ' fi'-T; 7-pcp" & Pvz. v-i-
■ ’ ■».; ; • ■■' v‘ " ■?■ ■' - X
: 1 1 9 •
; ii i . -1 . • 1 . ■ . < ■
69
All of the above mentioned deficiencies in the reaction
model are considered to be of minor importance compared to the
reactions which cause the product to have a brown color. It
is believed that this brown color in the product is caused
by consecutive aldoling of the aldehydes, and to a lesser de¬
gree ketones, found in the product. It is believed that this
aldoling is accompanied by dehydration, dehydrogenation, de-
carbonylation and/or decarboxylation. This results in the
formation of high molecular weight resins which contain rela¬
tively small amounts of oxygen. This is indicated by the high
oxygen accountabilities encountered in all runs. This alde¬
hyde condensation also accounts for the small amounts of 2-
methyl-2-pentenal found in the product, since it can readily
undergo further reaction with propionaldehyde.
The inability to determine the extent of these poly¬
merization reactions, due to lack of analysis, poses a serious
restriction on the analysis of the data since the amount of
hydrogen, carbon monoxide and carbon dioxide formed as a re¬
sult of these reactions cannot be determined. The magnitude
of this limitation will be dealt with in the following section.
Figure 12 shows that propylene is formed by two re¬
actions: from the dehydration of n-propanol and from the
decomposition of n-propyl propionate. It is difficult to
establish the magnitude of the dehydration of n-propanol. An
indication of the amount of dehydration which may occur is
'
.. ■ ' x 'i. . : . '. , ' ,a . .■a C-:j ■ ■
- •
'■ ,.. >■.!,
11 •' ' ©■ .. 1 r ; ■. ■■ ■ . :
c £ ;.f ass..m t . 1 Xm < .. •: v i %v 1 i •
0T. / r/o '...i::" P, j"; .pi/■ r , tJkjsVfl ,
; lit r.r'Voi
■ ■ ; stj.3
1 ‘ v-o c ;• v : i..
' ' : '' - ' ' .do i". k \d-y't
' " ' '} J' ’ . .. ,
5 J . |
.'..j:? iw ;. v 4.zzL M.£'lw to.
y ; - ;
■ . ■ ■
' : s j" - 1 O • ■! j ■ .c 1 ’ y; ,. ,:i; : '■ , : ,7 ? ■
.
70
given by the quantity of 2-pentene found in the product since
it is formed by the dehydration of 3-pentanol. Runs 7 to 12
indicate that the dehydration reactions are very temperature
dependent since both 2-pentene and propylene content increased
markedly with increasing temperature. The water content of
the product also increased appreciably with increasing tempera
ture; this gives further evidence that the dehydrogenation re¬
actions become more significant at higher temperatures. The
measured water content is not a very reliable indication of
the overall reaction sequence, not only because of the in¬
accuracy of water analysis, but also due to its disappearance
via reactions that can consume water, e.g. hydrolysis of n-
propyl propionate.
3. Aldol Versus Ester Route
The aldol route refers to the formation of diethyl
ketone from the aldol of propionaldehyde and 3-pentanol se¬
quentially as intermediates, while the ester route refers to
the route with n-propyl propionate, and in the case of hydro¬
lysis, propionic acid, as intermediates. To analyse the re¬
action model presented in Figure 12 on a quantitative basis
requires the determination of the contribution of each of the
above two routes to the diethyl ketone formation.
The reaction model shows that for each mole of n-
propanol reacting according to the ester route, one mole of
carbon dioxide is produced. One mole of carbon monoxide is
" ' i '
• . .
■■ .' .
■ ' ■ ' 3".: ' ,:v.
- • £
If i :
t, s ' t *;; : ■ ' 1/
' : - . .. : /
ft
* . . • ' ■ •
.
a ■ : . . i (
a
*t ' ■ :
71
formed for every mole of n-propanol reacting according to the
aldol route. The model also shows that three moles of hydro¬
gen are produced per mole of diethyl ketone formed by the
aldol route, while four moles of hydrogen are formed per mole
of diethyl ketone produced by the ester route, either by
hydrolysis or direct reaction of n-propyl propionate.
Table 12 shows the diethyl ketone yields by the ester
and aldol routes based on the assumption that all of the car¬
bon dioxide and carbon monoxide is formed only by those re¬
actions presented in Figure 12 (Case 1). Table 12 also shows
the predicted hydrogen yield for the two extreme cases.
Case 2: all the diethyl ketone is formed by the
aldol route.
Case 3: all the diethyl ketone is formed by the
ester route.
For comparison, the experimental diethyl ketone and
hydrogen yields are also included in Table 12.
The following relationships, based on the stoichio¬
metry of the reactions in Figure 12, were used to calculate
the values in Table 12.
Diethyl ketone by aldol route = [CO] - [3-pentanol]
- [2-pentene] - [ethane]
Diethyl ketone by ester route = [C0 2 ]
.
: w ; P . , ; . : ■ ■'
■
, ,■ .v,..rb - P '
. ■/ ' .CCJC
. ■ ■ ' V■ ■ ,v v .* . f .... ' ... ■ ' •' ■: ' ■ ;
. >■',: , ■..' : : <•, , ’0 ‘>;c}3 \to
■ I . ’ V : •" >, ... J ~XZi L
>1 :■ c<". ; 'vj ns s>. lux:..: ' ".'■■d
, .v i £. ' ■.. i r . ■
. ■ pc C : : i " ■">nc-
.... s d v fi P. i •■ ;'■-■■■ •/ •
i
■
. . ' .. .... ’ ;
72
Table 12
Comparison of Calculated to Experimental
Yields of Diethyl Ketone and Hydrogen
(quantities given in moles of product
per 100 moles of n-propanol feed)
Run
Diethyl Ketone
By Ester Route
1
Diethyl Ketone
By Aldol Route n
cn
H
11— 1
Total Calculated
Diethyl Ketone
Experimental
Diethyl Ketone
>
Hydrogen w
(Aldol only) w
to
n
Hydrogen gj
(Ester only)
U>
Experimental
Hydrogen
1
24.1
16.3
40.4
23.6
74.9
98.5
123.0
2
22.0
16.2
38.2
23.5
75.3
98.8
126.0
3
19.1
16.5
35.6
25.1
84.2
109.3
119.0
9
0.35
0.10
0.45
0.26
5.15
5.41
7.65
10
1.54
0.78
2.32
2.10
14.33
16.43
20.72
11
10.15
6.43
16.58
12.8
46.3
59.1
63.9
12
17.6
13.7
31.3
21.7
78.1
99.8
124.0
13
14.2
8.4
22.6
14.2
50.9
65.1
85.0
14
7.70
5.95
13.65
11.2
42.0
53.2
72.6
15
5.02
3.35
8.37
8.13
33.7
41.8
58.6
16
2.01
1.28
3.29
2.38
18.3
20.7
27.6
17
1.35
0.60
1.95
1.17
14.4
15.6
19.06
. . . .. ' . - ""
' ■ ' : ■ , .. "v• .i • ■ ■' '■ '■
..... ...... .. . ■ • • ■ •• • ■ • ••• • • •.. .
. ' . " . ' .". : ' ' ' ■ ' Vi ; ■'
q
.. i: is ■ j
■
■
:
■: a
... ■ •• • ■
' $ -.. ; '
• : xs
. Z
■ u l
73
Total calculated diethyl ketone = (diethyl ketone by aldol route)
+ (diethyl ketone by ester route)
Experimental diethyl ketone = [diethyl ketone]
+ 2 x {ethyl isopropyl ketone]
Hydrogen (Case 2) = 3 x [diethyl ketone] + 2 x [3-pentanol]
+ 2 x [2-pentene] + [propionaldehyde]
+ 2 x [n-propyl propionate] + [ethane]
Hydrogen (Case 3) = 4 x [diethyl ketone] + 2 x [3-pentanol]
+ 2 x [2-pentene] + [propionaldehyde]
+ 2 x [n-propyl propionate] + [ethane]
Experimental Hydrogen ■ [H 2 ]
where
[A] = moles of A in product per 100 moles of
n-propanol feed
The calculated results presented in Table 12 show that
the predicted diethyl ketone yield is always higher than the
experimental yield, and the calculated hydrogen yield, even
for the case of maximum hydrogen production (ester route), is
always low. These results agree with the earlier statement
that other unknown reactions occur, producing hydrogen, carbon
dioxide and carbon monoxide.
An attempt was made to calculate the amount of hydro¬
gen, carbon dioxide and carbon monoxide resulting from unknown
i'-'v'C ■■■ \ ©o;-: : r,/... ■ : :,i i£.?oE'
"-:.r
. •' •» £i"iirtafpigtoq^St
' t : : r : ^ -f
... . . ; \ ' v ■' ... »=- (:■ ' ^.tO ) '■•^.oxxbvli
' - • ■' 1 l. "-v u ■ ^ T 1 ' -
- ... '■• ; . v . ':■■■; :", v/
, c {] r:--p-'.:. . :• /.;£ ■:: yyy ©' x™
■ :'5 <-■ ;r*. ■
1: % sA ,.• ^EXiC:
. ; ■ . x : ,r l ■ ,j ; : ■:■ :
: '■ . f-'.:. ^ xv- /yy V; ry ©rH'
I ■ I ' .
' r t • ■ ■ 1 V .X: • . ’ r V ‘ ?*:' S IZtJ.
0 c fcio* to ti
. ' ; / 1 ;i ■: ■« ■ .
••••*» r.;,;' : ■ O-W 5 ■? •?« S’*W . : 7i Lr’
inn pA&Xiniftv ■ mxixso • t>&& Ghlxdlb rtodjm
74
reactions. The approach taken was to write overall hydrogen,
carbon and oxygen balances along with the stoichiometric equa¬
tions for the postulated reaction scheme. The number of equa¬
tions resulting from this procedure were less than the number
of unknowns and therefore could not be solved. The identifi¬
cation and analysis of the unknown products and the reactions
causing their formation is required to solve this problem.
Figure 12 shows ester and aldol route contributions
to the diethyl ketone yield based on overall carbon dioxide
and carbon monoxide formation. Besides the fact that the
overall diethyl ketone yield obtained this way is always exces
sive, the possibility of a disproporionation of the carbon
dioxide and carbon monoxide content due to the water-gas shift
reaction also exists. This causes this method of calculation,
to determine aldol and ester contribution, to be in error.
Table 13 lists the temperatures of various runs, the equilib¬
rium constant at the same temperatures(21) and the ratio of
experimental carbon dioxide and hydrogen to carbon monoxide
and water content in the product.
The results in Table 13 show that the water-gas
shift reaction does not go to equilibrium and the driving
force is always towards the formation of carbon monoxide and
water. An indication to what extent the water-gas shift re¬
action occurs may be obtained by investigating the amount of
carbon monoxide that forms or disappears when mixtures of
. ■ j,
.vjl.o:, b u.: r% r?ri<r yl&. t'.QOst&Z&i* * v?y£&cv ftns nod*
. ... f-vc "• v-. ■; •. v?> &iL y
. . ■' '■ ' * J
I •- '
■ -v .. . ■
'
' ■ • ■ i' '* : V : ; ' " • i
-■ ‘' • '
75
Table 13
Effect of Water-Gas Shift Reaction on
Carbon Monoxide and Carbon Dioxide Yields
Temp. Conditions Favor
Run °C K * R** Formation of
j>
1
465
6.44
202
CO
+ h 2 0
2
470
6.20
185
CO
+ h 2 0
3
464
6.50
44.3
CO
+ h 2 0
9
380 14.4
41.2
CO
+ h 2 0
10
419
9.79
12.5
CO
+ H o 0
2
11
459
6.77
26.8
CO
+ h 2 0
12
496
5.05
33.4
CO
+ H 2 0
13
463
6.56
48.2
CO
+ h 2 0
14
461
6.65
30.9
CO
+ h 2 0
15
466
6.38
31.4
CO
+ h 2 0
16
464
6.50
30.8
CO
+ H 2 0
17
462
6.59
42.7
CO
+ h 2 0
* K
P
= equilibrium
constant
[co 2 3 [h 2 ]
[CO][h 2 o]
** R = --- , as found in product
[CO][H 2 0]
.
..
' . ' T'J X' ,v. ,r;J. / " »
.
.
'
•
. , ... r ' )
»r .V , t
.
.
• •••-. .. ■ .wi ■ • ...
'■> { v ., ■
■
2 I :■
c~ •
?■
'» / ..
' . . ■ .. . . \ " . 4
- ..... •• -<• ■ " •
76
carbon dioxide, hydrogen, carbon monoxide and water are passed
over a base-treated Alundum catalyst. The effect of other
reactions, involving the above four compounds, on the water-
gas shift reaction is not easily obtained unless the rates
are known for these reactions.
B. Reaction Rates
1. Overall Rates of Formation'
For a heterogeneous catalytic reaction in a tubular
flow reaction, the rate of appearance of a compound, A, may
be defined by
= F
dN,
dw
dN,
d( S. V?
(27)
where
A
F
w
n a
S. V.
rate of formation of A (moles/hr-gm of catalyst)
feed rate (moles/hr)
weight of catalyst in reactor (gm)
moles of A per mole of feed
space velocity (moles of feed/hr-gm of catalyst)
If the experimental data is plotted as versus
1/(S.V.) the slope of such a graph, according to equation 27,
gives the rate of formation of A. The slope at different
values of the space velocity may be obtained by direct measure¬
ment or by a numerical differentiation. One such numerical
procedure calculates average rate values over an increment,
v- ■■■■■■' ;;/ ■ ’ , 5 5 1 :• .■ ■ ; ''v.r.^*..
■ •' ■ • ' v '■ ,■'> ! ■•.%■ V. ■ ■'■■ i..r /o
■
' ■ :'■■■'‘ • ' ■ : ' ^;•. ■. 1 i u .....
.' . v “ ■ ' "■
-<; . ' ,
yjrrf 1 at - 1 to s® 3 " IS a imt&
■ f* ■ - ? ;®.' ■ ... ■ - ' V-:*
. ' fi}' : A’"*
. , s
. ■■ ) : •...: 1 ,; • -. 1 ^c x ? : ;:■& .
osa) « %
. ." " . ■! s? , : \\ .... '•
■ rl •; «’ ; v.t
, ' . " ' ' . . ■ ;-.0-.r: . -. ■■•.fi
•. ; . ■ M i . ? .: . - r.'.-;jT ; ■.. ■; y$ ■■ T
■ ■ , ■ . . ■.
• ". ;-' ;i . b —v.v,: v■ , i-o Txi j Raot to &d: fcOVi^
• ■
77
A (-) , and the corresponding increment in N^, AN^.
gives
This
<rj
A
(28)
A AVE
The resulting average rates are plotted versus the
reciprocal of the space velocity. The instantaneous rates
are obtained by drawing a smooth curve through the average
rate values in such a manner that the areas under the two
plots are equal (see Figure V-l in Appendix V).
The overall rates presented in Figure 13 and Table
V-l are the result of using both of the above methods and as
can be seen from Figure V-l, good agreement between the two
procedures was obtained.
The rate curves of Figure 13 show the overall rates
of formation for eight compounds encountered in the dehydro¬
genation of n-propanol. In the following sections, factors
influencing the rates of heterogeneous catalytic reactions
and methods for correlating rate data will be discussed.
2. Factors Influencing Reaction Rates
Aside from the catalytic properties of the catalyst,
the rate of heterogeneous catalytic reactions and to a cer¬
tain degree the types of chemical reactions occurring may be
influenced by the following physical processes or properties
of the reaction system.
:: ' • • ■ •<... ■, . . 1 t ‘) ■'
..... ; ' ':-ri rum,■■■ :v rtsw# ;:*J: ->s.o /h olz.t
..... ■. ■ •
‘ ( y. v r?s :Vfia '. 10 $ '10
' ' ;.v£:. Oi.»-f ■ /C.
: . •" i | . '■ ' :
"■ : ■■■ f .o . V v ; I;;! &.;•:/ : 1 In v.
RATE X I0 4 (MOLES/HR.-GM. OF CATALYST)
78
FIGURE 13 « OVERALL RATES OF FORMATION
(SCALE FOR n-PROPANOL - Q ) (SCALE FOR HYDROGEN - (§) )
3 -
■" ' ••• &
: 1 \)
... .. r -w,
cm:
4'iv' ' ■ 0 . : ■ ■;
.
79
1. Diffusion of the reactants to the surface of the
catalyst.
2. Diffusion of the products from the surface to
the catalyst to the gas stream.
3. Surfaces, such as the walls of the reactor, may
display catalytic activity towards one or more of the com¬
pounds present in the reaction mixture.
4. Deposition of material on the catalyst's surface
may poison the catalyst.
The effect of these factors on the dehydrogenation of
n-propanol at the conditions encountered in this work will
now be discussed.
No conclusive evidence was obtained to determine
whether or not film diffusion, both of the reactants and the
products, was a limiting step in the reaction process. Runs
2 and 3 present evidence that film diffusion is not the rate¬
controlling step. These runs were carried out at similar
temperatures and space velocities, but different n-propanol
feed rates and hence different gas velocities. Similar n-
propanol conversions occurred in both cases, 94.8% and 93.6%
in Runs 2 and 3 respectively. The n-propanol feed rates of
these runs were the lowest encountered in any of the runs
and if diffusion is not important under those conditions, it
will definitely not be important at higher feed rates. The
results of these runs are not conclusive in determining the
I
:: a,. 1 1. .:Q
■ j ::
' ;.oo
: ■ • \ v' :
V;: ;.3 & V
d " ■, ■ 1 " ' ‘ "::o 1:
:.■ " ' i. '• *•••> ■ : *' *'
'
80
e ffect of film diffusion since the two runs were carried out
in different reactors. Although both reactors had approxi¬
mately the same inside diameter, other factors may influence
the reactions.
It is recommended that a series of experiments be
carried out to thoroughly investigate the effect of diffusion
from the gas stream to the catalyst's surface and the catalyst's
surface to the gas stream.
The catalytic effect of stainless steel towards the
decomposition of n-propanol was investigated in Runs 4 and 5.
Run 4 was carried out in the empty stainless steel reactor,
while Run 5 was done in the empty glass reactor. Run 5
showed a 2.8% conversion of n-propanol. This decomposition
was attributed to thermal rather than catalytic reactions.
A 6.7% n-propanol conversion was obtained in Run 4 and since
the temperature of Run 4 (440°C) was 20°C lower than the
temperature of Run 5 (460°C) it was concluded that the major¬
ity of the conversion in Run 4 was due to the catalytic ac¬
tivity of the stainless steel reactor walls.
Since the same products formed from n-propanol in
the presence of stainless steel also occurred if the Alundum
catalyst was used, the separation of the two catalytic ef¬
fects requires that an intensive study of this mixed cata¬
lyst system be undertaken. For this reason, no attempt was
made to determine separately for the effect that stainless
8
. - -: /" , xxix O'. •':• fiO.i '•x- j I ? x : * x r;: '
• " ; •, •. x.XiX ." ,. ■ : ' - : x x fxx xxx... 1
' ■ • x. ■: xx ■ x xxx.x ■ xx x,' ■ x- X-..". x ; x-' ;
, -vox : r, >%".i • ;i
' ., ■ x X iff--' ..
>- .v x • ■ pi;‘ xx . x : : - J-ji ' ' --'X V- - x - 1
XY-A.;-:-': ->~i OX ’ .X' X r, XX XXX KXt'l:
XXX X • X . X XX - 1 X
'
i . x tilt -'«esw lot
t ,, -■ ,:. : X XXXX'X XXXX X I .. . O O. ■ X V X< XX X i\ 0',i
.. x.x-- :< xx'. : x:- Cv^rxc
x. X - ■ ■ X . - ■■ '■ ' * X - : : 1 J xx . ;.-v ' .
' ! l '
• '■ X • :• • .X ■■=>
. ' ' 1 ■ '
- ' '
■
. ; • : , P,..... ' ,-vx XX- ... ■■ .■.,X:;,x ;.,xY :/ xoxx- xxxx
: : X. ■ xx.,. : : ' ' ... - .' x XI :. ■ xXx
' .
' • X.
81
steel has on the reaction products. The use of a reactor,
fabricated from a material inert towards n-propanol and any
subsequent dehydrogenation products, is recommended for further
study of this system.
Carbon was deposited on the Alundum catalyst during
its use and this deposition, in time, may cause deactivation
of the catalyst. To investigate the magnitude of this poison¬
ing, Runs 19 and 20 were carried out. As mentioned earlier,
the results of these two runs were in disagreement because
Run 19 showed a decline in catalytic activity while Run 20
showed no apparent decrease. Run 19 was done after the cata¬
lyst had been exposed to temperatures of 500°C and this may
account for the lower n-propanol conversion in Run 19. Since
Run 20 did not indicate a decline in activity over a period
of 50 hours of continuous use it was felt that the effect of
poisoning was negligible.
If the Alundum catalyst is to be employed for long
periods of time, perhaps several weeks, the effect of this
extended use on the catalytic activity should be investigated
further. If a decrease in activity is observed, methods of
reactivation should be examined.
3. Methods of Correlating Rate Data
Basically there are two approaches which are employed
in correlating kinetic rate data.
' . ■ 7 7
• ■ . •• :> ■" ■ 7 s 7 7- ■ . ss' .7 ; i ep 1:
• • . . /" sss
,777:1 &■} ■■■ 7-’ 'lie. V-l TS J Z:
. i ■ :
* ,:.:..77:7 1 ;■s. ^"SSs 7.7.1s • .7 I
• - 1 ■ 7 , ",7s, >
li:,-: ■ : ... " .. La :> sstsv 0£ bits <21 pnl
■ 7 7 ' ;> '.7 , ' 'J
7- ■ • . .7; ■ " •,:-b .7 j; v. 7 0 ! i 7 . 7“
■ L • 777; 7 ■' , - h • f rttxs gc & on b&w a tit
A" 7 • ‘ .7 S3*' ■ .A 7. 7t ?• 77S 'xpi-ii :7S: U b&fi S rfvl
: s
■ '.7' ■ ‘A " '
777 ■ 7 '7.07 7:77.777.7
7'
... ■
'
pe
7 " " -.. - • 7 ,. 7 ■ 7 < 7
.
- ' - '7 7 7.7'
■7, -. 7 ■ A
' 7 7 ,a >;, 7 ; .. •. • ■p -i 7 ,. < g xiv
■ _. 7 '''LL? ■; : 7 ' 7- ■ '777
S' , 7 ,7.7 \ 77 : " ., » ' -V "7 7 < ' 7 ■" f
- O 7 * S' , 7. 77 : , , ' '7
.'.77 ' a*... . 7 7 'Si 777.
82
1. The over-all reaction rate is correlated empirical¬
ly as a function of the operating conditions.
2. Rate equations are developed which describe a
rational mechanism by which the reactions may occur, and the
best fitting relationship is assumed to describe the reaction
mechanism.
The first approach has a limited scope of application
since it is not safe to extrapolate much beyond the range of
the experimental data. The usefulness of this type of analy¬
sis is also reduced by the fact that no insight into the steps
occurring during the reaction process is gained. As an il¬
lustration to this approach the rate of decomposition of n-
propanol is correlated empirically in the following section.
In the second approach, the rate data is fitted by a
model describing a plausible reaction mechanism. The most
widely used theory to describe the mechanism, involving chemi¬
cal steps only, by which heterogeneous catalytic reactions
occur consists of the three steps:
1. Adsorption of the reactants on the surface of
the catalyst.
2. Reaction on the surface of the catalyst
3. Desorption of the products from the surface of
the catalyst.
The above three steps assume that diffusion of the
reactants to the surface of the catalyst and diffusion of
■
• . ; ^ ^-uXlirui v:.ea:t ') <7 fi'O t'.J 2tt T?.’ £ ' C", '{,1
. ' f, ■' •
, , ' ■ 4 . : „ ' . ; L< ... '' i- f ■ *' •«*^
.... ...... , . ... , .. ' ' ' - ■' :
;■ ’ ' ■,.•. ..
a -J o.-ai:= :d .
.. ., ... ; .. ■. i *X vr .' ® ■ " ' '
• • . ;. ■ ; . / . . • ■ - V* I ':,gcr.i q
, :& . fcd,f
<</■) W’£ V a '> v °
■ . :. . ■' ' ' ' ' '
■ . .. : ... ;; '. . .. .. : ./ ..■ :? ■■■:
■ '■ .. : :j o: ' ■ . ■. '
- . . . : ;
. ' , ■■■ • : .vc : rw: • ■ ' a '■ v. 2 ; -
83
the products from the surface to the bulk gas stream are not
the rate-controlling steps.
This model assumes that the reactants are adsorbed
on active sites on the catalyst's surface where they react,
either by themselves, with other absorbed species on neigh¬
bouring sites or with components in the gas film adjacent to
the catalyst's surface. This reaction model is commonly
called the Langmuir-Hinshelwood model and Walas(33) gives an
excellent coverage of types of rate expressions resulting from
use of this model.
Although the approach in developing rate expressions
from the Langmuir-Hinshelwood model is straightforward, the
task involved in determining the best rate expressions for a
complex reaction system is immense. This is the case here,
since a large number of mechanisms may be postulated by which
the reactions can occur and the rate expressions from each of
these possibilities must be tested against the experimental
results. This determination of the best mechanism(5) usually
requires a large amount of kinetic data.
The size of the task involved is illustrated by the
work of Mezaki and Watson(18) who studied the catalytic oxi¬
dation of methane and tested 84 models of the Langmuir-
Hinshelwood type to determine the rate of oxidation of methane.
For a reaction system which contains several consecutive as
well as parallel reactions the complexity as well as the amount
: ' ■ -f ;j« ■ .. -sn. : -rd?
1 . . : ' 'V,.' * '• y i '■ *iC: ijV T'X t ~
- '-<■ ■ • &■ : & j . v • \ 1 '' ? ?mMfi vd m& 1 ©
-
V -*b nX -rly /i^diXA
a '■
' * S : ' X®IC
,b . : o-v* X'; 1 , '' \ \ ' . ’ .
.. ■ ' fc.:. ;•• voider
• £• ' ■!■:. ..•. - ■' ' V s- .
; • I. hi ; ' .. i ?,v , r'.
■ - ! V Xyn J ■: * j X ■ h J ; - ' ' 5V
84
of work involved increases considerably. It is also necessary
to obtain reaction rates for the individual reactions occurring
in the sequence. The number of possible reaction mechanisms
also increases significantly since an adsorbed compound in a
consecutive series of reactions may undergo a series of re¬
actions without the desorption of the intermediate products.
The number of models to be considered for each step
of the multiple reaction system can be reduced by experimentally
determining which of the steps, adsorption, surface reaction
or desorption, is rate controlling.
4. Correlation of Rate Data
a) Empirical Method
To illustrate this method the rate of decomposition
of n-propanol was fitted to a first order rate equation.
d(
dX.
A
1
S.V.
)
K(1 - X A )
(29)
where
X A = fractional conversion of n-propanol
K = rate constant (moles of feed/hr-gm of catalyst)
Rearranging and integrating equation 29 yields
A A dX. (S.V.) 1
/ - = K j d( -) (30)
0 (1 - x A ) 0 S.V.
1
= K -
-In(1 - X A )
(S.V.)
(31)
ft - ' ... -ft . ■ ft ft'ft ft
1
: - -. : : • ; ft ■' . - ft!',rftUv .. .
•'""■ft.' ft ft,"ft ft'"'. ft-ft ft ftfrOtvftftftSft to iftftfti ft-ft :&r/i ftv
>
V V ' Vj ft. - ft. ft. f ftftftilU ft .." ftftftft ft<9: fti
: ft 'ft - ,, ■. .'.ft' ft '.ft J ' •
• . : h .
. *■-' ft.;.ft; :
, ft; 't ft'ftftft. ft .ftJft'ftft J . ft'ft -.0 , &
-ftftftJ j . 1 ' (*’
■■ ■ - : r -ft;.- ft "ft; '.'ft
" ■ ' " . .' . ft ' ' ft '
■ ■ .. • ■ ; ' : ft.r"..' ft",.'.::; L. ■ ft: "..
"
■ ...., • ' ■ ■ ft-'- : , i.- '-ft -ft": i. ft,; • -■ ft ....:ftft. ftft'-ft.ft. % ftft'' ; '-::
85
To check the validity of equation 31 a plot of In (1 - X A )
versus 1/(S.V.) should yield a straight line. This plot is
shown in Figure 14.
Once a satisfactory rate expression has been obtained,
values of the rate constant at various temperatures can be
calculated from equation 31, which, upon rearrangement,becomes
K = -(S.V.) In (1 - X A ) (32)
The Arrhenius' equation describes a relation between tempera¬
ture and specific rate constants,
K = K e -*/ T <33)
o
where "K " and "a" are constants
o
T = temperature (°K)
Equation 33 may be written as
1
In K = In K - a(-) (34)
° T
Table 14 gives values of K versus 1/T calculated by
equation 32. Figure 15 shows that a plot of InK versus 1/T
can be approximated by a straight line. This indicates that
the catalytic reactions involving n-propanol follow the Ar¬
rhenius' equation.
Solving for the constants in the above equations, the
rate of disappearance of n-propanol is given by
r«r 1ft 6 -14,800/T,. v x
-r, = 6.55 x 10 e ' (1 - X a )
A A
(35)
•-I V
■ • ' . : '■ / -
. ' V 3 Vji • V V- 'JOl'iO
^ , . v : ; . • ; ■ ..
: / . ' >
, i m i r. I
,;m . i: ;.x ■■£ i>- ' ' ' u '■ “ &:>. ;
1 *
■
i . •; ■■ - • •; r,? ■ i
■ ■
tv' , it -Jttlq '" M- .SC aol^Mpe
. v: . ;• r: ; . ; ■; v ,? ......, — . .
' :■ . ::i\ . ■ . *\.:V ; '• ■ ■ " • r-: ■' . . vx■
' ' ■ ' -'f ■ , x'x >: x : xxx: v;': ■, i .
< : %
:
-PROPANOL CONVERSION)
86
I.O-
0.9-
0 . 8 -
0.7-
0 . 6 -
0.5-
0.4-
0.3-
0 . 2 -
- r 1 — --i---1-(- - — -1—
20 40 60 80 100
(SPACE VELOCITY ) _l
FIGURE 14 : TEST FOR FIRST-ORDER REACTION FOR
DECOMPOSITION OF n-PROPANOL
A-
K XIO 5 (MOLES OF FEED / HR.-GM. OF CATALYST)
87
FIGURE 15 ■ TEMPERATURE DEPENDENCY OF THE RATE OF
DECOMPOSITION FOR n-PROPANOL
/
‘OOOi
1006
oos
;
Or: ;•
• "a ■ •; .... >; ViU ■■ • V - r
51 '
- i . 0 . • r £1 • : ■ ! V- >
Run
7
8
9
10
11
12
13
14
15
16
88
Table 14
Effect of Temperature on Rate Constant
for Reaction of n-Propanol
Space
x
Velocity
2
10
Temp.
°K
K
x 10 2
1
Temp.(°K)
1.032
574
0.966
0.00414
1.742
1.040
612
0.981
0.0199
1.634
1.032
653
0.929
0.0762
1.531
1.043
692
0.808
0.223
1.445
1.035
732
0.446
0.837
1.366
1.025
769
0.029
3.63
1.300
1.163
736
0.314
1.35
1.359
1.825
734
0.480
1.36
1.362
2.818
739
0.572
1.57
1.353
7.60
737
0.784
1.85
1.357
■
P-
'■
. , v ...
SU .. :
■Ui X
tc*. c)
1C o
0 1 £ .
f-.: t
c : ■
89
b. Langmuir-Hinshelwood Model
Since it was not possible to obtain individual reaction
rates from the overall rates presented in Figure 13, no attempt
was made to apply the Langmuir-Hinshelwood approach to the de¬
hydrogenation reactions of n-propanol.
In the previous sections several sets of experiments
have been recommended which may make it possible to obtain a
reaction model for which rate data for each of the reaction
steps may be obtained. Once this is achieved the Langmuir-
Hinshelwood approach should be examined closely as a means
of correlating the rate data.
C. Yields
The yields of diethyl ketone from the dehydrogenation
of n-propanol at various operating conditions have been given
in Tables 4 and 5. If the intermediate compounds, propion-
aldehyde, n-propyl propionate and 3-pentanol, as well as the
unreacted n-propanol can be recovered from the reaction mix¬
ture and reacted again at the same conditions, higher yields
are possible.
Table 15 shows a comparison of the yields obtainable
at different operating conditions with and without the re-use
of intermediate products occurring in the diethyl ketone
formation. The yields as used in Table 15 are defined as
follows.
, v»v-J
©Id '
oo ’ : ■ -■ x .. :•
s J c :■ ,Id ao q;^ ".ojob\ q\\-q: -or: ;ooJ. *•. rb >{ Cqqo el.-00
' "■ ' ■; ■ '• AX A'if , ,-i ■ qq
1 ; .a >Som ' - tie, i o bo; . . . \ '■ oo aaac .■ o:
bo . .o b; borq sJ/yi ' ob bv y>:r : boboo: aci'Saa'ri
• b ■ ( " ■■ O'.. " o’-oo
t. ' ' 1 :
w. 'f : .i .a v ;. i .£ oo 1 ■ be
'
o:b ar.,:- a , . .0. >v;d
; .. ,'b; ' -q0o'
iO'w "i
0 Vis , iOftE:.; ooq-
■ qqooq ■ .ob q;. : 'boi>-
■
.■ &i : ' . .... v !,>, ; b: • \o 1 $
yo" ' ■: J r . .; - • : :/o. ? •. ■ •,II. :
>«*■• ' ■ £ . L r . r;>-, 0 ©
Run
3
9
10
11
12
13
14
15
16
17
90
Table 15
Comparison of Diethyl Ketone Yields
Space Velocity
Temp. (mole/hr-gm of catalyst) Yield 1 Yield 2 Yield 3
°C x 10 2 % % %
464
0.615
43.0
45.8
49.8
380
1.032
0.52
7.2
68.4
419
1.043
3.68
19.2
57.6
459
1.035
21.0
38.0
49.0
496
1.025
35.9
37.0
38.9
463
1.163
26.0
38.0
47.8
461
1. 825
18.8
36.2
50.5
466
2.818
13.6
31.8
51.6
464
7.60
4.1
18.9
69.0
462
19.17
2.1
13.4
80.0
■
'
8X5, S
91
Yield 1 — moles of diethyl ketone per 2 moles of n-propanol
feed with no n-propanol or intermediates recovered.
= 2 x (moles of diethyl ketone)/(mole of n-Propanol
feed)
Yield 2
= moles of diethyl ketone per 2 moles of n-propanol
feed with all of the n-propanol, but no intermedi¬
ates, recovered and recycled (Yield 2 is the same
as yields reported in Tables 4 and 5)
= 2 x (moles of diethyl ketone)/(mole of n-propanol
reacted)
Yield 3
= moles of diethyl ketone per 2 moles of n-propanol
feed with all unreacted n-propanol, propionaldehyde,
n-propyl propionate and 3-pentanol recovered and
returned to the reactor
= [(moles of propionaldehyde) + 2 x (moles of
diethyl ketone + 3-pentanol + n-propyl propionate)]/
(mole of n-propanol reacted)
Table 15 shows that yields (Yield 3) as high as 80%
(Run 17) were obtained. The values for Yield 3 show the effect
of operating conditions on the amount of side products formed.
At very low conversion and operating temperature at 460°C,
70% or more of the products formed were due to reactions oc¬
curring in the formation of diethyl ketone (Run 16 and 17).
As the n-propanol conversion increased with decreasing space
velocity at constant temperature (Runs 3, 13, 14 and 15) the
'
. ;,v.. J u • 'tfci-; ■/.•.■ 'H ' o it*: ' 0..: -
.
■ .. ■- . ■ J.,:--.- .6, *' ' <
: -T h -■ r - ^ - - ,i - ^ : -V -
ib : ■ -: •, : .. ' .%
■
>
y, . - I • .~o:
at.bfiB fci itv % t
92
amount of n-propanol that reacted to yield diethyl ketone and
intermediates in the diethyl ketone formation, reduced to 50%.
This indicates that the intermediates in the diethyl ketone
formation also undergo reactions that do not lead to the for¬
mation of diethyl ketone.
The runs at varying temperature and constant space
velocity show that even at low conversions (Run 9) more than
30% of the reacted n-propanol undergoes undesirable reactions.
Undesirable reactions are those which lead to production of
products other than, or those that ultimately lead, to diethyl
ketone. At a temperature of 496°C (Run 12) the amount of un¬
desirable product increased to 61%. This indicates that the
magnitude of the side reactions is very temperature dependent.
On the basis of these results it was concluded that
the maximum yield of diethyl ketone obtainable from the de¬
hydrogenation of n-propanol, using the catalyst employed in
this work is of the order of 50%. It is believed that the un¬
desirable products are mainly the result of thermal, dehydra¬
tion and aldoling reactions. The thermal reactions cannot
be eliminated but are relatively unimportant and probably do
not consume more than 5% of the reacting n-propanol (Run 5).
The aldol reactions are base-catalysed and may be partially
eliminated by reducing the sodium hydroxide concentration
introduced during the catalyst's preparation. However, this
would tend to increase the dehydration reactions since un-
■- . ' i'".v." 1 ; ..vwv- , , ■ tv '.a,-, :,r. . .nlo
Qfi. oi fteotrib&x -\nojifwtii oil £&'.• n’X v m$a±b&m®$nfr
■■..■■■ ■
v ::$:&£■%} ion oft i&li dais, noliisfftxoi
■■■.'■■■ ■
: > .c-• ■ ;j~: . ■ 1,0
nrv/ij v/ioto ■■■ ;-"t, /ku & i &yn.w- w »J nav*- ’ u®d» v^looifev
. .. • ; - > : .■. , ; ..v ' - v.-a. ■. • . ,GnO'A;■ oz jxv- rx eri:v v : iOC
f v; i ■ - ntv ; . . 0:10 s ‘ ow ' ,vv 1
IvtiS^ib hf>®i '-i1: ri£tj $j&ii$. B&Qt&t.' to iit&tfS
"■At ? ; . t"'.'.: ' . " "(1 . O' ! ■ . OV -t.v £,• .vt r'.; V 7
:. ; o'.v: '■ -io ■ ; s ■ io.ynom‘ iJdi tv
■ . . ■■ i ,■ ’>1 : v iti .O f " v
i :V :6W . V L. 5 g ; jlfii.. v . taSfJ ., .7 llO '
“.vc - ■ i tnfvv- V. ir wit", / ; : o. ;v;. Ot v ■
•: . r , ■ -pr tl ' . pm :piq~tv >:>• aii£vf>pa ft id
- ■ v. aOv, v : .:; . .. v o -tv"?-:;, ani : 0v >.v too ■ v,t ;iv
•• ■■ ■ , A ., .A,.' 1 "t " " V a V : fh . ■ V ix,.:i>:v; tv.:, t ' tv Vt-iv,.;---.;Otlvot, t'vvvv :
: AA y, ' , ) V ./vi . Z: V V t .. V
t .:■■• ; ; . tv v ' .A ■ - .1: * V V. ■ A ; : A ■ ,r_:„
t . , • ‘ • ■ - -
.. v ..v fei - i: -
. . vc.-;- v- v..av: t. t a ,-ti ; . .• _v' : Diivtv $ali:ub ovovr ooicvt
93
treated Alundum is a dehydration catalyst. A study to deter¬
mine the effect of sodium hydroxide concentration on the
catalyst's selectivity is necessary to determine the optimum
sodium hydroxide concentration. Komarewsky and Coley(13)
found that for sodium hydroxide treated chromia an optimum
sodium hydroxide concentration existed.
94
VI. CONCLUSIONS
A. Catalyst and Reaction Sequence
1. The basic alumina-silica catalyst (sodium hydrox¬
ide treated Alundum LA-617) employed in this work was active
towards reactions of n-propanol. A conversion of 97.1% of
— 0
the feed was obtained at a space velocity of 1.025 x 10
moles of n-propanol/hr-gm of catalyst and a temperature of
496°C.
2. Although n-propanol conversion increased with in¬
creasing temperature and decreasing space velocity the maxi¬
mum diethyl ketone yield is obtained at temperatures between
_2
460° and 465°C and space velocities less than 1.0 x 10
gm-moles/hr-gm of catalyst. For the above range of conditions
a diethyl ketone yield of 40 to 45% is obtainable.
3. The reaction products obtained and the analysis
of the data show that diethyl ketone is formed by two major
routes, i.e. via aldol condensation and via the Tischenko
condensation. The presence of water in the product also in¬
dicates the possibility of hydrolysis of n-propyl propionate
to propionic acid followed by the formation of diethyl ketone
from propionic acid. The relative extent of reaction occurr¬
ing by the two competing routes remains unresolved.
4. The dehydration of 3-pentanol to 2-pentene in¬
creased markedly with increasing temperature. The amount of
■
» J j . J ,;a.s^ ■
•J ■ :>
: : J ■' VS 3WC :• XOC’&qoyq^a d^QdSiii c .|f» .£
' 1 '■ . (■:% "
! ’ ■
' ■; 1 -v, ■X lb
• • ■" V: :
\'OW
. ! H ■ Vo
. -f t .... i
; v. ..■••• ■
sv... ■■■,. -//•
.
v ' ■ : ■■'■;■ c ■ : •. . . v ’■' <;•■■
. IT
2-pe.ntene formed per mole of n-propanol reacted was 0.012 5
and 0.0392 moles at 419°C and 496°C respectively. An increase
in propylene content was also noted with increasing tempera¬
ture and it was concluded that this increase is due to the
dehydration of n-propanol to propylene rather than an increase
in propylene formed from the reaction of n-propyl propionate.
5. The reaction model postulated for the dehydrogen¬
ation of n-propanol shows the main reaction sequences by
which diethyl ketone is formed as well as several side re¬
actions that occur, but other reactions occur as well and the
chemistry of these has not yet been determined. Due to the
lack of a more complete reaction model it was not possible to
determine reaction rates for individual reaction steps from
the overall rates of formation.
B. Equipment
1. The lead-bismuth eutectic as the heat transfer
media, together with the automatic temperature controller,
resulted in satisfactory isothermal reaction control. The
average maximum temperature difference encountered between
various locations in the catalyst bed was found to be 6°C.
2. The pressure control obtained by manual control
of a valve was not adequate since pressure fluctuations of
±1 psi were observed. It is believed that these fluctuations
are due to the drop-wise vaporization of the feed.
■ 1 ~ ' \'C oo '•
- ' . ... : . ' ,
SO • .. .. - h: X l . .. • 0 L 1 .0 . ' . IJ :
: ■ g . f ' '''.■■■• ■ J{j i '
..: : ) r.' ! 'v :• -oho ;oo::‘? orrery: -:,oo.yv y- , |
...v ; • o . o . ‘ i ho o:r... 7 ■„ o o.„r.l
■•v. : ... ro ,:■;■■■ ,h: ye--'roro - r - D,u yo-rJ'y
• y .v h.eyyvyr.' oor i. ox. t ' o i .so r i r , h r v;i.t o rh
t ■ s to . . " .' i
■■-■■a ' o.vo ' ' srs .... o :r ; - :i c 'Osydd
■ y ' :■ '. r><o - -oen: s: ! s Us/os:; soon r lo 0o‘h,I
■ . O.SO j. : o j". .;.;os 0 ' - 0 ,;Vr 0 :. o iX
. ; S,, : : ;■ -I 0 .... y • y ;. ,f , . ■ V, : ■. .
sna/rfqx.ytdo
.... - •. . ■ ' • o . ■ 0 : * SO-. V -O';.- s/h;
If. .} '3~J.yS2iVr.Bq.nC ' 'J oh: Mil'S
Ml ,i "Mi .. . ■ $ i 0 : ’J i
ho? :o :>on® $or.%? eoc j&v&qm&l ojjnj fern ag&xa &
a .
■ ■ . . .0 : : ' • -J "I X S' 0 O 'CO ^.'i’SKy::' ■ df.V
70 ‘o' \ . 'a xxmUk rdsop'yhB soo? *:.:.■<■ I tv - "he
; :
. -• - ■ i - ■■ « 01000 ' ■ ■’ ho o hr o. ; h
96
VII. RECOMMENDATIONS
The following list of suggestions may aid future in¬
vestigators studying the dehydrogenation of n-alcohols to
symmetrical ketones in a fixed bed reactor.
A. Reactor
A reactor meeting the following requirements would be
useful in the study:
1. The construction material should be such that it
does not promote the decomposition of alcohols and subsequent
dehydrogenation compounds. The use of a glass reactor in the
present cerrobase heating bath presents difficulties and
therefore, the investigation of some alternate metal alloys
or high temperature coatings is recommended.
2. The feed system to the reactor should be such
that gas-liquid mixtures could be introduced to the reactor.
This arrangement would make it possible to dilute the feed
stream with inert gases and thereby attain very high space
velocities.
3. The reactor should be easily removable from the
heater and the port used to introduce the catalyst should be
at the top of the reactor so that it does not become coated
with cerrobase. This would facilitate changing of catalyst
charge, at present, a very difficult task.
: . : r:‘ .
•• ■ ' ’ i& : >v;. i
.
97
B . Experiments
The following experimental work would yield further
insight into the chemical reactions occurring..
1. The use of anhydrous n-propyl propionate as feed
in one run, and water-saturated ester in another would help
to establish the hydrolysis route if a difference in propylene
formation is observed. These runs should be carried out at
high space velocities.
2. The use of propionaldol (if obtainable) as a feed
would prove or disprove the assumption that the majority of
the unaccounted materials result from propionaldol.
3. The use of carbon dioxide, carbon monoxide, hydro¬
gen and water mixtures, of known compositions, as feed would
give the magnitude of the water-gas shift reaction under the
reaction conditions.
4. The importance of a more complete product analy¬
sis cannot be over emphasized. This point will be discussed
more thoroughly in Appendix III.
'
say vT.
: i'S ■
- i . J ..
■
v: .
^ - ;..yrr ■ ,
U.. i til >:
, -
“ ’ ' '• ' ' tf
' . .... : :.:
q ’i&i. v j .
& U * ■ J ' ' : 7 . ■ .
» 1 .& v: ' - ..
T l: ■ ' r ■; . ■ ■
98
BIBLIOGRAPHY
1. Bevington, J.C., "The Polymerization of Aldehydes,"
Quarterly Reviews, 6_, 141 (1952).
2. Church, J.M. and Joshi, H.K., "Acetaldehyde by Dehydro¬
genation of Ethyl Alcohol," Ind. Eng. Chem.,
£3, 1804 (1951).
3. Day, A.R. and Joullie, M.M., "Organic Chemistry," Toronto,
D. Van Nostrand Co. Inc., 1960, Chapter 19.
4. Engelder, C.J., "The Decomposition of Ethyl Alcohol,"
J. Phy. Chem., 21, 676 (1917).
5. Hansen, K.T., "Dehydrogenation of n-Propanol and Propion-
aldehyde with Chromia Catalysts," M.Sc. Thesis,
U. of Alberta, Edmonton, Alberta, June 1960.
6. Hinshelwood, C.N. and Hutchinson, W.K., "A Homogeneous
Unimolecular Reaction - The Thermal Decomposition
of Acetone in the Gaseous State," Roy. Soc. Proc.,
A 111 .245 (1926).
7. Hinshelwood, C.N. and Thompson, H.W., "An Apparently Uni¬
molecular Reaction - The Homogeneous Decomposi¬
tion of Gaseous Propionic Aldehyde, " Roy. Soc.
Proc., A 113 , 221 (1926).
8. Hurd, C.D., "The Pyrolysis of Carbon Compounds," New
York, The Chemical Catalog Co. Inc., 1929.
9. Hurd, C.D. and Meinert, R.N., "The Pyrolysis of Propylene,"
J. Am. Chem. Soc., 52_ r 4978 (1930).
10. Ipatieff, V.N., "Catalytic Reactions at High Pressure and
Temperatures," New York, The Macmillan Company,
1936, p. 441 - 451.
11. Ibid., p. 125.
12. Kegan, M.J., Sobolew, I.A. and Lubarsky, G.D., "Folgen-
reaktionen bei heterogener Katalyse, III Mitteil:
Der Mechanismus der Bildung von Acetone aus
Acetaldehyd and Wasser," Ber., 68_, 1140 (1935).
13. Komarewsky, V.I. and Coley, J.R. , "Catalytic Synthesis
of Ketones," Advances in Catalysis and Related
Subjects, £, 207 (1956).
)£'■ A A 10 : 'ISI '
i. :• " ■ ’C CCi . V " . l’(
,
■
■, .mw ,-iM X
; :x •. • i-lT- ' 1 " T ■ 1 c c iOYi v
1 Ct; -v ‘v .. ' cmO re* -c.a iO
H r>. . . h ■ . A ; ■ ;>v .
3 ) 8 il -■ •. :o ' r . 3 o*ofl. / luo*>.ta-it
-v " . , ..' ... ■, . CO i
, { ■ i .o xs<: a i soxf
- ^c.. j .Tv f.. ' d vl ■ • . . c 1 jccioi f. ,^1
; V: : M. ©rf Vi." V t- ■: :•
. 'Ted; , J 5
vcc AA , wo.' vv. cl ,t o , -c c
v • x ■ .. : " “v: i f ov - :. / •■;> i
, w -$4; '■ - t sL
. vr:v :: . j ■ ,, . o •
i D . V 0 V ■ T ' ' V clV ■■■
99
14. Lazier, W.A. and Adkins, H., "The Formation of Carbon
Dioxide from Alcohols," J. of Phy. Chem., 30,
895 (1926).
15. Marek, L.F. and Hahn, D.A., "The Catalytic Oxidation of
Organic Compounds in the Vapour Phase," New York,
The Chemical Catalog Co. Inc., 1939, Chapter II.
16. Messner, A.E., Rosie, D.M. and Argabright, P.A., "Corre¬
lation of Thermal Conductivity Cell Response with
Molecular Weight and Structure," Anal. Chem., 31 ,
230 (1959).
17. Mezaki, R. and Watson, C.C., "Catalytic Oxidation of
Methane," Ind. Eng. Chem. Process Design Develop.,
5, 62 (1966).
18. Morrison, R.T. and Boyd, R.N., "Organic Chemistry,"
Boston, Allyn and Bacon Inc., 1959, p. 486.
19. Noller, C.R., "Chemistry of Organic Compounds," Phila¬
delphia, W.B. Saunders Company, 1957, p. 209.
20. Pines, H. and Haag, W.O., "Alumina: Catalyst and Sup¬
port. I. Alumina, its Intrinsic Acidity and
Catalytic Activity," J. Am. Chem. Soc., 82 ,
2471 (1960).
21. Perry, J.H., "Chemical Engineers' Handbook," 4th ed.,
Toronto, McGraw-Hill Book Company Inc., 1963,
p. 23-45.
22. "Physical and Thermodynamic Properties of Elements and
Compounds," Pamphlet No. GC245R2-5-858, Chemical
Products Division of Chemtron Corporation,
Louisville, Kentucky, 1958.
23. Rice, R.O. and Rise, K.K., "The Aliphatic Free Radical,"
Baltimore, The John Hopkins Press, 1935.
24. Rosie, D.M. and Grob, R.L., "Thermal Conductivity Be¬
havior," Anal. Chem., 2_9 1263 (1957).
25. Royals, E.E., "Advanced Organic Chemistry," New York,
Prentice Hall Inc., 1954, p. 750.
26. Scheidt, F.M., "Vapor-Phase Aldol Condensation over
Heterogeneous Catalysts," J. of Catalysis, _3,
372 (1964) .
100
27. Schneider, V. and Frohlich, P.K., "Mechanism of Formation
of Aromatics from Lower Paraffins," Ind. Eng.
Chem., 23, 1405 (1931).
28. Taylor, II.S., "Fourth Report of the Committee on Contact
Catalysis," J. Phy. Chem., 30, 145 (1926).
29. "Tentative Test for Water Using Karl Fischer Reagents;
ASTM Designation: E 203-62T," ASTM Standards
1962. Supplement Part 7, Philadelphia, Ameri¬
can Society for Testing and Materials, 1962,
p. 387.
30. Tollefson, E.L., personal communication, Chemcell (1963)
Limited, Edmonton, Alberta.
31. Vasudeva, K., "Reactions of n-Propanol on Alumina Sup¬
ported Chromia Catalyst," M.Sc. Thesis, U.
of Alberta, Edmonton, Alberta, April 1963.
32. Vasudeva, K., "Vapor Phase Reactions of n-Propanol on
Solid Catalysts," Ph.D. Thesis, U. of Alberta,
Edmonton, Alberta, September, 1965.
33. Walas, S.M., "Reaction Kinetics for Chemical Engineers,"
Toronto, McGraw-Hill Book Company Inc., 1959,
Chapter 7.
'
■
> - j - ■ : .. i..i
■ ,-u • . ■ : i" <
y : \ ' ^ •• ■ ' ■
APPENDIX I
THERMOCOUPLE EVALUATION
. ■ ' ...
:
1-2
A. Accura cy of Ther mo co up le s
Since no reference thermocouple was available for
calibration of the thermocouples used in the reactor a rela¬
tive calibration was used. This was done by placing the eight
thermocouples into an electric furnace and comparing the tem¬
peratures recorded by the thermocouples at various furnace
temperatures. To minimize variations in the thermocouple tip
temperatures due to their location in the furnace, and radia¬
tion and convection heat transfer, the eight thermocouples
were placed in a bundle with their tips touching and wrapped
with asbestos insulation.
Table 1-1 shows the results of this calibration.
Table 1- 1
Thermocou pl e Calibration
Temperatures
Recorded
at Various Oven
Setting;
Thermocouple
l.Yo *
Setting
No. 1
Setting
No. 2
Setting
No. 3
Setting
No. 4
TC 1
126.3
268.8
337 .0
501.8
TC2
126.6
2 39 .9
338.1
502 .6
VC 3
126.8
271.0
338.8
503.2
VC 4
126 .7
269.6
338.2
503.1
TC 5
126.4
269 .7
337.3
501.1
VC 6
127.3
270.4
338.4
503.8
TC7
126.0
268.5
336.6
501 .8
TC8
126.3
270.0
337 .8
502 .4
. • • s ■-■■■
■
-
■ • ■ , • ' ; • 3 - :
'
J Q.'
. S‘> :
1-3
Since no absolute reference was available it can only
be concluded that the variation between the eight thermocouple
readings is not larger than * 2°C. It should also be noted
that thermocouple TC6 probably reads 1°C too high while ther¬
mocouple TC7 reads 1°C too low.
B. Errors in Thermocouples Due to Heat Transfer
Figure 1-1 shows the construction of the thermocouples
as found in the reactor. The thermocouples were bent at the
centerline so that longer thermocouples could be used in order
to minimize the difference between the thermocouple tip tem¬
perature (T ) and the surrounding gas temperature (T_) due to
P G
heat transfer along the thermocouple.
Figure 1-1
'
3-jXc - o - ■' ; ft - i ' '■ t n o i .1 o; r ,/. j rtc
:
■
.
1-4
The heat balance for the thermocouple at steady state
conditions is given by
Q + Q = Q + Q, (1-1)
gr c r k ' '
where
Qg r = rate of heat flow due to gas radiation
= rate of heat flow due to convection
Q r = rate of heat flow due to radiation
= rate of heat flow due to conduction
Since the thermocouples were located in a bed of
catalyst particles both radiation contributions will be neg¬
lected .
Equation 1-1 with this assumption becomes
Q k - Q c = 0 (1-2)
The conduction term in Equation 1-2 is not dependent
on whether the thermocouple is parallel or perpendicular to
the gas flow. The convective heat transfer coefficient, h ,
V
to the thermocouple well depends whether the gas flow is
perpendicular or parallel to the thermocouple well. In the
case where the thermocouple is located in a packed bed this
difference may be partly eliminated.
In the following calculations a constant value for h c
is assumed. This value was estimated to be 30 Btu/hr-ft °F .
This value of li is an estimated value since no data for heat
c
.
m
' ; S'-') < • .*v •: ) ' r ; ® ' , ■
' ! ; .i -;:5f . ,• ■ - . ' " •: .■ ' - 4 . • 1 ' '• { • '"V. L; 1;
. . ' . IS " S '"; x ..,0 ■■ -■
.. J i f; ■ ii vq&i ' l * ■ i■ < c
1 ::■■■■• -" ' ' ,.:•■■■ ’>■ c 2 l . - j-.v, • ’*1 ■ -*•■' !
r '&• :"r,: *i ■- >,i a v $>r\. s.kvs'w
f , I i! : M. - 5 • :.• > X 2 i ' to It 3 ■ ■■■'•■r o',. $j-Lj al
: ' V
1-5
transfer to small cylinders inside a packed bed were available.
The value was arrived at by consulting "Perrys' Chemical
Engineers' Handbook" fourth edition. Section 10, and "Heat
Transfer" Volume II by Jakob, Chapter 42.
Figure 1-2
With the above assumptions the energy balance over a
thermocouple length increment. Ax, becomes
d dT
— (ks —) Ax - h PAx (T - T) =0 (1-3)
dx dx C G
where
k = thermal conductivity of thermocouple well
= 12.1 Btu/hr.ft.°F
P = perimeter of thermocouple well
= 1.638 x 10~ 2 ft.
= cross sectional area of thermocouple well,
(metal part only)
= 5.0 x 10" 6 ft 2
s
• : C ' :
■ :: ■ ■■ i' ' ■ '
'
,
• ...v 1
.
S - £ %
,'y. ■ ■- ■; ' ' r T J ' ' ■ X ■ ■ X"' 1 - ■' ? ■
. n J- ■■;,yV?:X J X qx VX <Ji '1 -iftl-
xxx,. y :
Th :•>
l-J:> Kb
>■;. V . . ■; xv qX .. / r J :-ff'huOj i ' X:Xor' ,■
X : x-w x.uy.-xi ■■ X\‘x.; XxijXi to,
; : '. .. ■ ■': v x xx . j ; . " ■ xox:,
1-6
T = temperature of thermocouple well at point x
Equation 1-3 may be written as
d 2 e
B 2 0 = 0
(1-4)
where
T - T,
B =
h P
c
ks
The general solution of equation 1-4 is
0 = C 1 e Bx + C 2 e Bx
(1-5)
To evaluate the two constants, and two boundary
conditions are required.
1. at x = 0 ;
2 . at x = L;
T = T
w
0=6
w
de
dx
= 0
Boundary condition 2 is an approximation since it
neglects the heat transfer at the tip of the thermocouple.
Solving for constants and C 2 by using the boundary
conditions and solving for the gas temperature (T-) in terms
of the temperature at the tip of the thermocouple (T ) results
P
m
P _
w
cosh BL
( 1 - 6 )
■ v.:. t '' fc
'-yf:.'y.ry: v b-l os .rt ■ ob
r-i ■
:
• ' . no/:, // ■ iD noi : . /OS 1 £.1 i> nov : 0
oi at ■■
■
.
iXb
j ' - : . ' .
'[ - •£
;; on
.' ■ : r%&$muo£
1 -■■/ o nx/l pitiv lo?>
■- / ■■ - o. - .. ■1: s c 3: ' - ■ . - Lc bi 1 / &r?o " v ■ s
■' ■ o . . .. . ■ : : S : ; • '■ 3tiz : . oJIi 3 m e U :i<)
J
JS d&OO
Sample Calculation for Thermocouple Error
The largest temperature difference between the wall
and the thermocouple tip occurred in Run 17. For an illustra¬
tion the temperatures of this run will be used.
T = 1350 °R
w
T = 1323°R
P
The value of B is
30 x 1.638 x 10
12.1 x 5.0 x 10
= 28.5 ft
L = 0.96 inch = 0.08 ft.
Substituting these values into equation 1-6
6
E
e
T,
cosh 2.28
0.203
w w G
Solving the above equation for T
T^ = 1316 °R
Therefore,
T - T_ = 7°F = 4°C
P G
For the case where T = 461°C and T = 466°C the dif-
p w
ference between T ar>d T^ is 1.3°C. Since this is a small
p G
error, no corrections were made to the temperatures recorded
V . • t /. 7 f'' a :v vioitf
J. , 0 '■ . ; i ' • ■
: ; ,/ ; / "'H ? .■..•> I? -
■-■ T
T' • ’ >. j-,,. „ ,• ,-!»•» nnrni -.- t -• ,|,v, r ■ . 7.-\" ‘
H'". “ ..
.... ... . r »,
'
. V. ;; - r ■ ■ -.V .r ; ' ■- " ■
1,-8
by the thermocouples.
To determine whether it is worthwhile to install the
thermocouples with a bend at the centerline, or only project
them horizontally to the centerline of the reactor, the fol¬
lowing comparison in thermocouples errors is made.
For the thermocouple projected only to the center
L is 0.0368 ft. Considering the temperatures encountered in
Run 17 and substituting the values into equation 1-6
e
JBL =
e
w
1
- = 0.595
cosh 1.11
Solving for T
P
T
G
T - = 22 °C
P G
The above calculation shows that the construction of
the thermocouples as employed in this work greatly reduces
the error in thermocouple temperature reading.
. . . a . .. y ■ i D «£» ; ..‘C ' ^
. .-.ft-i ' 5«t* . 3
a
fll"
.... •" ' ->• ■ • ’ ' ■ l '-
APPENDIX II
SAMPLE TEMPERATURE PROFILES
1 -".L
: . '
• ■' U-f 1 ‘ . i
II-2
Table II-l
Temperature History for Run No. 4
Time
(min.)
TCI
TC2
TC3
TC4
TC5
TC6
TC7
TC8
TC9
0
461
462
462
461
457
469
467
439
197
2
461
462
462
462
458
469
466
439
4
461
462
462
461
458
469
467
438
6
461
462
462
462
458
469
467
439
8
461
462
462
462
458
469
466
438
10
461
462
462
462
457
469
467
439
201
12
461
462
462
462
456
469
467
440
14
461
462
461
461
456
469
457
441
16
461
461
461
461
456
469
463
441
18
461
461
461
461
456
469
467
441
20
461
461
461
461
456
469
467
441
194
22
461
461
461
461
456
468
467
440
24
461
461
461
461
456
469
467
440
26
461
461
461
461
456
468
467
439
28
461
461
461
461
456
469
467
438
30
461
461
461
461
457
468
467
437
19 3
32
461
461
462
462
458
469
467
438
34
461
461
463
462
458
468
466
438
36
461
462
463
463
458
468
466
438
38
462
463
46 3
463
459
469
467
438
40
462
463
463
463
457
469
466
439
193
42
461
463
463
462
457
469
466
440
44
461
463
463
463
467
468
466
440
46
461
463
463
463
456
468
466
441
48
461
463
463
461
456
468
466
441
194
50
461
462
461
461
456
468
466
441
52
461
461
461
461
456
468
466
440
54
461
461
461
461
457
468
466
440
56
461
461
461
461
458
468
466
440
58
461
461
461
461
459
468
467
439
60
461
461
461
461
458
468
466
440
19 3
NOTE:
Thermocouples
TCI
to TC4
are located
in catalyst
bed.
Average catalyst temperature = 461°C
Difference between wall and centerline temperature = 5°C
Axial temperature gradients - essentially zero
•: ; ...
r;yf
•’ '.rll ■ V*.' •
■
.1 :.
' r-
-
:■ >
Xd£
.
■
J&ft
. r&x.
.v
'
£ 3*
■M'
■•■d*
( j a&‘
■; v
' "I*
I
Qh-&
'.a£ ' X-3&
i ;
i:
8dp
Id*
■
3dP ■
* v
sm
X^'-i*
II- 3
Table II-2
Temperature History for Run No. 17
Time
(min.)
TCI
TC2
TC3
TC4
TC5
TC6
TC7
TC8
TC9
0
461
464
477
476
476
478
477
468
200
2
461
464
477
476
476
478
477
468
4
460
464
477
476
47 5
478
478
467
6
460
463
477
47 5
47 5
478
477
467
8
460
463
477
476
47 5
478
477
467
10
460
463
477
476
476
478
477
468
201
12
460
464
477
476
476
478
477
468
14
460
463
477
476
476
478
477
468
16
460
463
477
476
476
478
477
467
18
460
463
477
476
476
478
477
468
20
460
463
477
476
476
478
477
467
200
NOTE: Only thermocouples TCI and TC2 located in catalyst bed.
Average catalyst temperature = 462°C
Difference between wall and centerline temperature =
Axial temperature gradient = l°C/inch of bed height
15°C
•. -I X
n-
V \ a
. r ...
* *
r.. -
rt&
: T'4
■ ' *
:vc;--
rti
t
d i
iv,.
-
Odt*
' Vfr
V V'v
c £
Eat
^ p-
7
DM-
■
■
: :0 ■ t-. I SOT \> r : /- ■' ' , v\ ;
. ,q .■ ::r :.l ■ ■ >&•.. ■ :
APPENDIX III
PRODUCT ANALYSIS
t
III-2
A. Sample Analysis
To illustrate the separations obtained by gas chromato¬
graphic analysis, sample chromatograms for liquid and gas analy
sis are shown in Figures III-l, III-2 and III-3. The condi¬
tions under which these chromatograms were obtained have been
described in the body of the report.
Table III-l shows the retention times for the identified
products in the liquid analysis. The relative retention times
for each compound compared to the retention time for n-propanol
The time for air to elute was subtracted from all the retention
times for computation of the relative retention times.
The column and the operating conditions used to ob¬
tain the values shown in Table III-l have been described in
the section on product analysis.
B. Recommendations for Analysis
By examining the chromatogram for the liquid analysis.
Figure III-l,it can be seen that a large number of unidentified
products are present. Some of these unidentified products
may be
II
/
CH^CH-C
3 2 \
OH
1. the hemiacetal
0
ch 2 ch 2 ch 3
Figure III-l
Chromatogram for Liquid Product Analysis
Propionaldehyde
CN
X
X
X
-Propanol
Ethyl Isopropyl Ketone
Ethyl Isopropyl Ketone
I
I*
III -3
L'con on Celite Colurm
03
X
'
Charcoal Column
III-6
Table III-l
Retention Times of Liquid Products
Product
Retention Time Relative Retention
(min.) Time
Air
2.1
2-Pentene
7.0 0.244
Propionaldehyde
9.5 0.368
Methyl Ethyl Ketone
16.4 0.710
n-Propanol
22.3 1.000
Diethyl Ketone
24.4 1.11
Ethyl Isopropyl Ketone 28.7 1.32
3-Pentanol
32.7 1.53
n-Propyl Propionate
35.0 1.63
2-Methyl-3-Pentenal
44.8 2.12
3-Heptanone
50.2 2.39
.
yv
III-7
,o-ch 2 ch 2 ch 3
2. the acetal CH 3 CH 2 CH
o-ch 2 ch 2 ch 3
3. the aldol
OH
I
CH 3 CH 2 CH
Besides checking the liquid product for the above
compounds it is also recommended to analyse for propionic acid
on a gas chromatographic column used for organic acids.
C. Gas Chromatographic and Orsat Gas Analysis
The gas product from 3 runs, Runs 10, 11 and 12,
were anlysed for hydrogen, carbon monoxide, carbon dioxide and
unsaturated hydrocarbons by orsat as well as by gas chromato¬
graphy. Table III-2 shows the results obtained by the two
methods of analysis.
Although the quantitative agreement between the gas
chromatographic and orsat results is not good, the orsat analy¬
sis gives evidence that the peaks on the gas chromatogram were
identified correctly. It is believed that the gas chromato¬
graphic analysis is more reliable since the author, who did
the analysis, was not familiar with orsat gas analysis.
.
a - •'
'
.. VK>'
r;
■■ T '■ ' . x. /• ■
III-8
Table III-2
Comparison of Gas Analysis
(Using gas chromatography and orsat analysis)
Run
No. 10
Run
No. 11
Run
No. 12
Method of Analysis
G.C. *
Orsat
G.C.
Orsat
G.C.
Orsat
Compound (analysis
by volume
per cent)
Hydrogen
84.0
82.8
74.4
77.7
71.0
68.4
Carbon Monoxide
5.1
5.4
9.4
8.7
12.2
12.2
Carbon Dioxide
6.3
6.2
11.8
9.3
10.0
8.7
Unsaturated
Hydrocarbons**
2.5
2.2
2.6
3.2
3.0
4.6
* G.C. - gas chromatographic analysis
** unsaturated hydrocarbon content reported for gas chromato
graphic analysis is the propylene content of the gas.
...
• _ <5 - 2 fp'>
o ::s ^y-yxxjot&flityzsixf. sja-v ;uus”!
*
■ ■
APPENDIX IV
OPERATING CONDITIONS AND ANALYSIS
FOR ALL EXPERIMENTAL RUNS
IV-2
Table IV-1
Operating Conditions
03
—.
0
•
cAO
0
c
4-t
•H
£
0
03
0
O
o
0
0
0
0
c
-P 0
i — 1
•H
+J
0
O »w
0
W
•H
p
r—t
3
£
a 4
0
0
Pd
0
03 4-1
-P
—-
5-1
o <N
m
0 0
o o
3
rH
o
p-i
P
P o
0
-P
0
rH
O
rH
fit 0
03 rH
•
u
u
0
>
0
rH
0 \
0
0
3
P
X
x:
•H
03 0
P 0
-p
0
0
0
w
-p
P
•P £
Pt 0
u
r cs
0
P-.
□
tn
0
P \
rH
C
0
o
0
e
0
C
-P
tr £
U> 0
2
CD
o
u
0
P,
0
0
°h tr>
0 £
Pd
Pd
In
Pa
EH
CO
PI
s
pi
U —
1
Glass
n-Propanol
13.6
464
0.769
60
93
38.1
176.5
2
Glass
n-Propanol
13.6
471
0.697
60
94
37.8
175.8
3
S.S. *
n-Propanol
30.8
464
0.615
60
95
40.7
167.1
4
S.S.
n-Propanol
28.0
440
-
60
99
59.8
6.7
5
Glass
n-Propanol
13.6
460
-
60
95
59.9
1.9
6
Glass
PrH***
13.6
480
-
30
95
57.9
4.0
n
i
S.S.
n-Propanol
31.8
301
1.032
60
99
59.9
0.4
8
S.S.
n-Propanol
31.8
339
1.040
60
98
59.9
1.9
9
S.S.
n-Propanol
32.2
380
1.032
60
98
59.6
8.4
10
S.S.
n-Propanol
32.4
419
1.043
60
94
58.0
24.6
11
S.S.
n-Propanol
32.4
459
1.035
60
97
50.3
85.9
12
S.S.
n-Propanol
32.4
496
1.025
60
96
32.5
174.9
13
S.S.
n-Propanol
32.2
463
1.163
60
96
47.7
114.8
14
S.S.
n-Propanol
32.1
461
1.825
60
96
52.6
89.5
15
S.S.
n-Propanol
32.6
466
2.818
20
99
54.8
69.5
16
S.S.
n-Propanol
32.6
464
7.60
60
100
57.3
32.7
17
S.S.
n-Propanol
32.6
462
19.17
20
101
58.3
22.4
18
Glass
3-Pentanol
13.6
462
0.72**
30
91
84.6
102.0
19
S.S.
n-Propanol
31.8
465
0.601
30
101
53.6
131.9
* S.S. - stainless steel reactor
** Space Velocity for Run 18 in gm of 3-pentanol fed per
hr-gm of catalyst
*** Propionaldehyde
, ■ I ■
. !
'
Of
3. If
oc; ; :
s . s •'
k £
i ■ •!.
■ £ k - j ,.v
. -- >
• a
- Xf S
• v .0
: K . In L r w L
IV-3
Table IV-2
Liquid Product Analyses
(weight per cent)
c
<D
Run No.
2-Pentene
Propionaldehyde
Methyl Ethyl
Ketone
n-Propanol
Diethyl Ketone
Ethyl Isopropyl
Ketone
3-Pentanol
n-Propyl
Propionate
■P
c
o
cu
i
CM
1
H
>1
x:
4J
s
i
<N
3-Heptanone
Water
1
1.0
1.59
4.18
7.00
44.13
5.30
0.30
0.30
0.51
6.58
0.4
2
0.83
1. 86
4.53
8.32
44.25
5.29
1.00
0.52
Trace
4.52
0.4
3
2.6
2.89
4.40
9.49
45.43
4.40
2.31
0.80
Trace
6.44
1.1
4
-
2.50
0.48
93. 61
-
-
-
0.70
-
-
0.6
5
0.18
1.16
0.48
97.16
-
-
-
-
-
-
Trace
6
-
98.17
-
-
-
-
-
-
-
-
-
7
-
1.36
0.04
99.80
-
-
-
0.12
-
-
Trace
8
-
0.88
0.19
98.27
-
-
-
0.29
-
-
0.06
9
-
2.55
0.31
93.52
0.38
-
-
1.67
Trace
-
0.15
10
0.29
4.90
0.72
83.59
2.73
0.22
0.17
2.28
0.17
0.28
0.6
11
0.85
4.96
1.60
53.18
17.95
2.31
0. 87
0.99
0.25
1.00
0.9
12
8.17
1.43
4.30
5.30
47.50
5.69
1.22
0.50
0.84
5.65
1.7
13
0.68
5.57
2.31
39.50
23.44
1.30
1.11
1.09
0.41
2.99
1.0
14
0.40
5.71
1.92
54.75
15.40
1.65
0.75
1.59
0.43
2.62
0.9
15
0.38
6.65
1.26
62.63
10.69
1.20
0.39
1.84
0.70
1.87
0.75
16
0.01
8.21
0.74
82.09
3.06
0.30
0.09
2.67
0.10
0.26
0.3
17
-
8.20
0.54
86.39
1.58
0.09
-
2.47
Trace
0.10
0.3
18
14.98
-
2.48
-
76.54
1.58
0.84
-
-
-
1.3
19
0.77
4.08
2.82
23.22
25.59
3.62
0.40
0.40
0.40
3.22
1.5
A " f. • V • ••***!••
•. • • . . ... ..>•
• ,1 • .1
>... *;
■>: .1
■
■.W,Q
•3k.:o,
■1
Wf*
■-
0‘8.. $
■
.
.€.£ 0
O'. .•
a .0
".
:
' .,o
OK 0
■: - . O...C
£ <2 , : 5
8£ u 0
„ ■
i.: .
31
: .•
3f:,. M
■k: o
IV-4
Table IV-3
Gas Product Analyses
(volume per cent)
<u
Q)
Run No.
Hydrogen
Carbon Monoxii
Carbon Dioxidi
Propylene
Propane
Ethane
Methane
1
70.1
10.5
13.7
3.6
0.40
0.85
1.0
2
70.6
10.6
12.5
3.0
0.34
0.86
1.0
3
71.0
11.4
12.1
3.1
0.36
0.72
1.1
4
74.5
4.5
6.0
6.0
-
6.0
3.0
5
73.6
5.3
-
10.5
5.3
5.3
-
6
-
50.0
-
-
-
50.0
-
7
100.0
-
-
-
-
-
-
8
100.0
-
-
-
-
-
-
9
91.0
1.5
4.2
1.1
1.5
0.36
0.36
10
84.0
5.1
6.3
2.5
1.0
0.61
0.41
11
74.4
9.4
11.8
2.6
0.48
0.61
0.71
12
71.0
12.2
10.0
3.0
0.88
1.9
1.0
13
74.0
8.9
12.4
2.9
0.35
0.61
0.88
14
81.2
7.8
8.6
1.4
0.30
0.30
0.45
15
84.2
5.9
7.2
1.7
0.29
0.27
0.43
16
84.3
5.0
6.1
2.1
0.40
0.91
1.13
17
85.5
3.4
6.5
2.1
0.40
0.67
1.4
18
96.4
1.15
2.45
-
-
-
-
19
71.2
10.7
12.7
3.1
0.96
0.53
0.96
NOTE
: Traces of ethylene, butane and
butenes
were also
detected in most of the gas products.
I ,sx
*> , I .
i
ox
***
£ .
, Cl
: .&/ ■
• ■ z'n id . . .• -o- :
, .2$xiuh 'Jlc ■ :■ i • (H io ;; ao/n ax l t
APPENDIX V
OVERALL RATES OF FORMATION DATA
r-
I
V-2
Table V-l
Rates of Formation
a)
rH
-P
I
id
c
d)
0
dJ
-P
•H
d)
<u
•H
>i
a
A
'd
•H
O
si
0
0
•H
X
0
o
u
-p
X
0
rH
rH
nd
CU
rH
d>
o
e
d)
o
rH
0
•H
o
>
e
fd
rH
c
CJ
Q
s
fd
e
>i
fd
d>
rH
d) -
a
0
a
-P
>.
C
c
o
o
•H
o
Cl
0
Si
O
0
fd
u
a
u
d)
u
-P
,Q
J3
a
cu
0
Oi
cu
'd
d)
P
S-i
CO
1
u
i
1
>i
•rH
fd
<d
c
c
ro
H
Q
U
U
Rates of Formation - (moles/hr-gm of catalyst) x 10 -
5
-170
o
•
o
4.2
-
-
20.6
15.3
19.0
10
-112
-4.4
0.5
-
170
19.8
15.0
19.0
15
- 98
-6.6
-2.6
1.5
145
19.2
14.7
19.0
20
- 86
-8.1
-2.6
1.2
126
18.7
14.5
19.0
30
- 73
-9.0
-1.7
1.0
104
17.5
13.9
19.0
40
- 62
-6.5
-1.2
0.8
92
16.5
13.6
19.0
50
- 56
-5.2
-1.0
0.7
69
15.6
13.3
18.0
60
- 50
-4.3
-0.8
0.6
60
14.1
13.0
17.0
80
- 42
-3.3
-0.5
0.5
46
10.4
12.4
16.0
100
- 36
-2.6
-0.3
0.4
42
8.4
11.5
11.4
120
- 30
-2.3
-0.2
0.3
41
6.9
10.5
10.9
140
- 23
-2.2
-0.2
0.2
40
5.4
9.0
10.1
\ ■■ > ' ovi; • - ;; ■:
> v X
-
.
■ - >.
i -
0,1 -
03
RATE OF FORMATION (XIO
V-2
20 40 60 80 100 120 140
(SPACE VELOCITY )' 1
FIGURE V-l : ILLUSTRATION OF PROCEDURE USED TO OBTAIN
INSTANTANEOUS RATES. DIETHYL KETONE USED
AS EXAMPLE.
