Skip to main content

Full text of "Dehydrogenation of n-propanol on an Alundum catalyst"

See other formats




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.