Improved Procedures for the Vapor Phase Nitration of Propane

Citation preview

P U R D U E U N IV E R S IT Y

T H IS IS TO C E R T IF Y T H A T T H E T H E S IS P R E P A R E D U N D E R MY S U P E R V IS IO N

by

James Veith Hewett

E N T ITL ED

IMPROVED PROCEDURES FOR THE VAPOR PHASE

NITRATION OF PROFANE

C O M P L IE S W IT H T H E U N IV E R SIT Y R E G U L A T IO N S O N G R A D U A TIO N T H E S E S

AN D IS A P PR O V E D BY M E A S F U L F IL L IN G T H IS P A R T O F T H E R E Q U IR E M E N T S

FO R THE DEGREE OF

Doctor of Philosophy

P R O F E S S O H IN C H A R G E O F T H E S IS

H

ead o f

S

chool or

D

epa rtm en t

■Tb V-^VTO T H E L IB R A R IA N :---_ rV-IS T H IS T H E S IS msSS&S TO B E R E G A R D E D A S C O N FID E N T IA L .

PROFB8SOH Hi OHABGB GRAD, SCHOOL FORM 6—3.49—1M

IMPROVED PROCEDURES FOR THE VAPOR PHASE NITRATION OF PROPANE A Thesis Submitted to the Faculty of Purdue University by James Veith Hewett In Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy February, 1950

ProQuest Number: 27712235

All rights reserved INFORMATION TO ALL USERS The quality of this reproduction is d e p e n d e n t upon the quality of the copy subm itted. In the unlikely e v e n t that the a u thor did not send a c o m p le te m anuscript and there are missing pages, these will be noted. Also, if m aterial had to be rem oved, a n o te will ind ica te the deletion.

uest ProQuest 27712235 Published by ProQuest LLC (2019). C opyright of the Dissertation is held by the Author. All rights reserved. This work is protected against unauthorized copying under Title 17, United States C o d e M icroform Edition © ProQuest LLC. ProQuest LLC. 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, Ml 48106 - 1346

ACKNOWLEDGMENT

This work was begun under Dr. H. B, Hass, but the major portion of the work including all of that done with nitric acid, was directed by Dr. G. B. Bachman. Both of these men provided wise counsel and original ideas during the time they served.

Dr. Bachman has been extremely

helpful in reading and suggesting corrections concerning the writing of the Thesis. This research was supported in part by funds provided by Commercial Solvents Corporation.

IMPROVED PROCEDURES FOR THE VAPOR PHASE NITRATION OF PROPANE WITH NITRIC ACID* By Dr. G. B. Bachman and J . V. Hewett Department of Chemistry and Purdue Research Foundation Purdue University, Lafayette, Indiana AN ABSTRACT SUMMARY The increase in conversion in the nitric acid nitration of propane resulting from the use of oxygen is obtained at the expense of diminished yields of nitro paraffins based on propane. While the use of high surface to volume ratio reactors is partially able to offset this condition the absolute values of yields obtained are still low.

The use of steam in combination with high surface

and oxygen increases the conversion based on nitric acid somewhat and further improves the yield based upon propane * Steam diluent shifts the distribution of by-products in the nitration reaction away from propylene and towards the formation of aldehydes. In nitrations with steam the use of oxygen beyond the quantity indicated by a propane to oxygen ratio of 5*1 is undesirable.

The use of bromine to

the extent of 0.0 15 mole per mole of nitric acid increases the conversion based upon nitric acid and almost doubles the yield based on propane. Bromine in this concentration ♦Contains material from Mr. J. V. Hewett*s doctoral thesis.

ii. makes the use of large amounts of oxygen undesirable and thus far the best results have been obtained with 0.0 15 mole of bromine per mole of nitric acid and a propane/oxygen ratio of 8.2.

Under these conditions a conversion of 48%

and a yield of 55% have been obtained. IUTHODUOTIOU In an investigation of the effects of various gases on the vapor phase nitration of propane, Hass and Alexander (3 ) discovered that the addition of oxygen to the reaction mixture caused an increase in the conversion of nitric acid to nitro paraffins.

The production of

oxygenated by-products was noted but no determination was made of the type of compounds produced or of the amounts of each formed. In an investigation of the effects of varied surface to volume ratios on the vapor phase nitration of butane, Addison (1) discovered that high ratios of surface to volume in the nitration reactor reduced the formation of by-products.

This work was extended to include propane

in several runs which showed that a surface to volume ratio of 28 would be desirable for further investigation. The main problem remaining in the nitration of propane with nitric acid was to find some way in which the conversion could be kept high and the yield on propane improved.

The rising price of propane emphasizes the

importance of this problem.

ill. APPARATUS The apparatus used was exactly the same as that used by Addison (l) in making his runs with a surface to volume ratio of 23.

The size of some of the metering devices

was changed to permit flows out of the range of those em­ ployed previously. PRODUCT ANALYSIS The methods of product analysis were the same as those used by the authors in their work in the nitration of propane with nitrogen dioxide and oxygen (4). NITRATION OF PROPANE WITH HIQH SURFACE AND OXYGEN A temperature of 425°0. and a contact time of 1.6 - 1.7 sec. were used.

This temperature was somewhat

higher than had been used by Alexander (3 ), but was necessitated by the poor external heat transfer of a reactor made of big tubing (22 mm. O.D.) and packed with glass helices in order to give the desired surface to volume ratio.

It should be kept in mind that further improvements

over the results obtained would probably result if the heat transfer between the reactor and the molten salt bath were improved. A series of runs was made using varying amounts of oxygen and a reactor with a calculated 23/1 surface to volume ratio.

This series showed that the use of increased

surface prevents a rapid fall in yield as the amount of oxygen in the feed is increased.

The operating conditions

and the results obtained are summarized in Table I.

This

iv.

Table I Effect of Varied Oxygen Feed Rates 1

2

4

Temperature, °C«

k-25

425

425

Contact time, sec.

1 .5

1.5

1-7

Run

2g

23

C-^/Op

21.6

10.0

Cj/ERO?

11.0

12.5

8/7 Mole Ratios

28 7.2 20.

Op/HRO^

0 .5 1

1.25

2.8(

HpO/HRO^

1.42

1.42

1.4;

Conversion ^ (N)

41

51

53

Yields $

HN02

39.2

34.0

3 4 .3

co2

4.1

2 .9

1 .7

2 1 .7

3 0 .3

3 3 .9

Ethylene

6.9

7 .0

2 .5

CO

9.3

7 .3

9 .7

18.9

18.0

1 7 .7

Propylene

Carbonyl Compounds

V

series of runs showed that, while the yield can "be prevented from falling off seriously with a reasonable increase in the amount of oxygen used, the yield was still not as good as was desired. NITRATIONS WITH STEAM, HIGH SURFACE AND OXYGEN The use of steam as a diluent in connection with high surface was suggested by the results of Addison’s (l) work with butane.

Thus three factors shown to be independently

desirable in the nitration reaction were brought together for the first time in the nitration process.

The data listed in

Table II show the effect of varying the nitric acid feed rate while the propane and oxygen feed rates are kept constant. This series of runs was necessary to determine if the addition of steam would affect the optimum balance between temperature and contact time verified for operation without steam.

It

may be concluded that the use of steam does not shift this balance.

Under optimum conditions steam increases the

conversion somewhat and the yield to a significant extent. The distribution of by-products is shifted away from the formation of propylene and towards the formation of more aldehydes. EFFECT OF LOW PROPANE OXYGEN MOLE RATIOS WITH STEAM AND HIGH SURFACE The shift in the distribution of by-products brought about by the use of steam with oxygen and high surface suggested the possibility that relatively large amounts of oxygen and high surface might be employed advantageously to

vi« Table II Effect of Varied Nitric Acid Feed Rates

8

5

9

6

Temperature, °0.(Preheat)

340

265

350

265

Temperature, °C.(Reactor)

425

425

425

425

Contact time, sec*

1-3

1.4

1 .8

1.9

Run

8/7

28

Mole Ratios

28

C^/Og

5.1

5.1

5.1

Cj /HNO-j

8 .6

9.7

1 6 .8

o2 /hno.

1.7

1.9

3-3

HgO/HUO

28 5.1 20

3.9

15

15

15

15

35

54

62

54

31-7

37.3

37-6

36 .6

0 .6

1 .3

0 .8

2 .6

2 8 .7

27 .0

26*6

1 5 .8

Ethylene

1 .2

3.2

0 .8

1 1 .8

CO

7.3

8 .7

9.3

6.9

30.5

2 1 .9

2 5 .I

2 6 .2

Conversion $ Yields $

28

(H)

RHOg co2 Propylene

Carbonyl Compounds

vii# increase the conversion to nitro paraffins and to increase the sum of the yield of nitro paraffins and aldehydes. This was thought desirable since aldehydes are readily separated from the recycle stream and have substantial commercial value.

The results of experiments in which the propane/oxygen

mole ratios were changed from 5*1 to in Table III.

to 3*2 are given

The desired effect was not observed.

Con­

versions and yields to nitro paraffins were not improved by these changes.

The formation of aldehydes was increased

slightly and then decreased as the oxygen feed was increased. This fact coupled with the rise in the formation of carbon monoxide suggest that the aldehydes are formed only to decompose in the reactor. NITRATIONS WITH STEAM, HIGH SURFACE AND BROMINE The use of bromine in the nitration reaction was tried as a method of generating hydrocarbon free radicals which would be relatively free of the high by-product producing properties of oxygen.

DeVries (2) has measured

the thermal dissociation of bromine and found it to be very slight at these temperatures.

However, Kistiakowsky and

Artsdalen (5) have proposed that this dissociation is a step in the thermal bromination of methane at temperatures lower than that used in the nitration experiments.

The explanation

of these facts is undoubtedly that the dissociation is small but that only a few bromine atoms are needed to start the process which is propably a chain reaction.

The effect of

viii. Table III Effect of Increased Oxygen Feed 9

10

11

Temperature, °0*

425

4-25

425

Contact time, sec»

l.g

1 .2

1*9

Run

s /v Mole Ratios

22

O-ZOg

22

22

5*1

4.0

3*2

1 6 .2

14.1

1 2 .1

o2 /hho.

3*3

3*5

3*9

HgO/HHO,

15*0

1 5 .0

1 5 .J

Conversion ^ (K)

62

57

52

Yields f»

37*6

3 6 .1

35*2

0 .2

1 .2

2 .2

2 6 .6

2 6 .0

2 2 .0

Ethylene

0 .2

0 *6

1 .2

CO

9*3

2 .1

14.0

25 .l

27.9

25 .5

Oj/HNO

RN0„ CVJ

8

Propylene

Carbonyl Compounds

bromine is best explained by the following reactions: (1)

Brg

(2)

C-^Hg -j-

2Br • Br*

HBr

O^Hy

Thus the bromine forms hydrocarbon free radicals (3)

HBr — |—

[jdJ

Br *

In reaction ( 3 ) the symbol oxidizing agent.

HO*

-j—

[o j

stands for any

Nitric acid, nitrogen dioxide, and

oxygen are all present in the nitration mixture.

There is

little evidence that reation (3 ) does form the HO* radical but the known properties of HBr can leave little doubt that it would be oxidized. ( 3)

The bromine atom produced in reaction

is free to repeat reaction (2) so that in passing through

the reactor the number of free radicals which one bromine atom can generate is potentially large.

The mechanism of

the nitration reaction from this point on has been discussed by Addison (l) and consists of the reaction of NOg with a hydrocarbon free radical to produce either a nitro paraffin or an alkyl nitrite. thermally.

The alkyl nitrites are very instable

They decompose to form nitric oxide, an aldehyde,

and a lower hydrocarbon free radical which itself may be nitrated. The results of the use of bromine without oxygen are given in Table IV.

Note that the conversion has been

increased from 23$ to almost 22% and that the yield based on propane has been almost doubled.

The actual molar quantities

listed allow a comparison of the amounts of each by-product produced.

Table IV The Effect of Bromine on the Nitration Reaction Without Oxygen Run

14-

15

Temperature, °C*

4-23

4-23

Contact time, sec.

1 .2

1 .9

s /v Mole Ratios

Conversion $

22

22

20.4-

2 2 .2

HgO/HNO

1 5 .0

1 5 .0

Br2 /HN0,

0

Cy'HNO,

(N)

Yields on Propane

Moles Recovered

0 .0 ]

2 3 .O

27*7

RNOg

27.4-

5 0 .0

002

2.7

none

Propylene

24-.3

Ethylene

2 3 .4.

1 5 .0

CO

1 2 .6

19.5

Carbonyl Compounds

9 .5

1 1 .0

RNO^

0 . 04-25

o .04-59

0.0119

none

Propylene

0.0 35 2

O.OO39

Ethylene

O .0517

0.0 19 6

CO

O.O556

0.0 51 0

Carbonyl Compounds

O.O34.9

0.024-0

4-.5

xi. NITRATIONS WITH STEAM, HIOH SURFACE, BROMINE AND OXYGEN The use of large amounts of oxygen with bromine is undesirable.

This is shown in Table V where two similar

runs, one with and the other without bromine, are compared• The use of intermediate amounts of oxygen with bromine produces good conversions and very high yields compared to the highest yields previously reported or found in the research.

The results of two runs (numbers 17 and

18) made with intermediate amounts of oxygen are found in Table VI together with the results of the other two runs made using bromine.

This allows a study of the effect of

varying the amount of oxygen.

The effect of oxygen on

conversion and yield is shown in Figure 1.

Figure II shows

the variations in the production of all the principle products of the reaction with oxygen feed.

The use of

bromine in the nitration reaction has opened up many possibilities of improving the conversions and particularly the yields based upon hydrocarbons in vapor phase nitrations. The full results of its use must await further investigation. ACKNOWLEDGMENT The authors wish to acknowledge the financial assistance of Commercial Solvents Corporation and the Purdue Research Foundation in carrying out this research.

xii. Table V Effect of Using Bromine with Oxygen in the Nitration of Propane with Nitric Acid Hun Temperature, °0.

9

16

423

423

Contact time, sec.

1.8

S/V Mole Ratios

28 C^/Og

1.8 28

5*1

5*1

O^/ENO

1 6 .8

I6 .5

o2 /hno3

3 .3

3*1

HgO/KNO^

1 5 .0

1 5 .0

Brg/HNO^

0 6 1 .7

Conversion % (N) Yields on Propane

RNOg

COg

3 7 .6

46.0 33 .2

0.8

0.5

26 .6

1 6 .1

Ethylene

0 .8

7 .6

00

9*3

13*3

Carbonyl Compounds

2 5 .1

29.4

RNOg

O.I37 S

0.1030

COg

0.0081

0.0042

Propylene

0.0928

0.0473

Ethylene

0.0040

0.0334

00

0 .0 96 8

0.1170

Carbonyl Compounds

0.2190

0.2170

Propylene

Moles Recovered

0.0 1 5

xiii. Table VI The Effect of Varying Oxygen Feed in

15

18

17

16

Temperature, °C-

423

423

423

423

Contact time, sec.

1.9

1.7

1 .5

1 .8

Run

28

S/V Mole Ratios

28

28

28

-

11.5

8 .2

5.1

o3/mro3

2 2 .8

10.6

9*9

1 6 .5

Og/HNO,

0

1.2

3*1

HgO/HNO,

15.0

15.0

1 5 .0

C^/Og

0.015

Br2 /HN0,

0 .9 2 1 5 .O 0.015

0.0 15

0.0 1 5

Conversion $ (N)

27.7

4 3 .2

4-7 .7

46.0

Yields on Propane RNOg

5 0 .0

57.0

5 5 .5

33*2

none

0 .2

none

0 .5

4.5

7.3

9-7

1 6 .1

Ethylene

1 5 .0

7*3

4.1

7*6

CO

19.5

1 .0

3 .6

13*3

Carbonyl Compounds

1 1 .0

26*6

2 7 .O

2 9 .4

COg Propylene

Moles Recovered*

RNOg COg

0.0459 none

0.1371

O .1765

0.1030

0.0017

none

0.0042

Propylene

0.0039

0.0164

0.0294

0.0473

Ethylene

0 .0 19 6

0.0262

0.0184

0.0334

CO

0.0510

0.0065

O.O331

0.1170

Carbonyl Compounds

0.0240

0.1490

0.2445

0.2170

♦All adjusted to same run time.

xiv. Figure 1

EFFECT OF V A R Y IN G in

n it r a t io n

w it h

Aa5°C

O

Pr

o pan e

O X Y G E N FEED b r o m in e

0.0/5 Y fttp

Q

l 7±S£C,

n it r o g e n

c o n v e r s io n

50

40

ZO

10

o

/

MOLES

z

3

O^/m OLE P N 0 3

4

XV* Figure 2

v a r ia t io n

in

m olls

W IT H OXYGON F L I P

OF p r o d u c t WITH R R O M /NE

• ETHYLENE. a co ■ ALDEHYDES

O /? /V q □ co2 X PROPYLENE .70

JO

{♦}------z

2

J

M O L E S Oz / M O L L H N C 3

xvi. LITERATURE CITED (1) Addison, L* M., A Study of the Vapor Phase Nitration of Butane with Emphasis on the Effect of Adding Oxygen, Ph. D. Thesis, Purdue University (1950)• (2) DeVries, Thos., and Rodebush, W. H., The Thermal Dissociation of Iodine and Bromine, J. Am. Chem. Soc Ï2 , 656-65 (1927 )* (3 ) Hass, H* B., and Alexander, Loyld, Oxygen Induced Vapor Phase Nitration of Paraffins, Ind. Eng. Ohem., 4-1, 2266-70 (1949). (4) Hewett, J. V., Improved Methods for the Vapor Phase Nitration of Propane, Ph. D. Thesis, Purdue University (1950 ). ( 5 ) Kistiakowsky, G. B., and. Van Artsdalen, E. R., Bromination of Hydrocarbons I. Photochemical and Thermal Bromination of Methane and Methyl Bromine, J. Ohem. Phys., 12, 469-73 (19#).

EFFECT OF OXYGEN ON THE VAPOR PHASE NITRATION OF PROPANE WITH NITROGEN DIOXIDE* By Dr. G. B. Bachman and J. V. Hewett Department of Chemistry and Purdue Research Foundation Purdue University, Lafayette, Indiana AN ABSTRACT SUMMARY The conversion obtainable in the vapor phase nitration of propane with nitrogen dioxide at practical contact times may be substantially increased by the addition of oxygen to the reaction mixture.

The optimum conversion

obtainable at two minutes contact time with oxygen is appreciably higher than at contact times up to and including fourteen minutes without oxygen.

The use of oxygen improves

the yield of nitro paraffins based on propane over that obtainable without oxygen by increasing the conversion to nitro paraffins more than the conversion to by-products. The yield based on propane is much higher using nitrogen dioxide as the nitrating agent than that obtained using nitric acid as the nitrating agent and present nitration techniques• In addition to the high yields based on propane, nitrations using nitrogen dioxide may be carried out so as to achieve high yields based on the nitrating agent. INTRODUCTION Extensive work on the vapor phase nitration of propane with nitrogen dioxide was carried out by Dorsky (3 ). *Contains material from Mr. J . V. Hewett fs doctoral thesis

ii. He found that nitrations with nitrogen dioxide are favored by long contact times and low temperatures compared to nitration with nitric acid.

Alexander (2) studied the

nitration of propane with nitric acid extensively and showed that the presence of oxygen increased the conversions materially. He also made two preliminary runs using nitrogen dioxide and oxygen and showed that an increase in conversion resulted.

There remained to be investigated

the full extent of the beneficial effects of oxygen on the nitration of propane with nitrogen dioxide.

Neither of the

above investigators determined the yields of their products based on propane• Addison (1), however, showed that the use of oxygen in the nitration of butane with nitric acid tended to decrease the yields of nitro paraffins and increase the yields of by-products based on the hydrocarbon. If the same applied to nitrations with nitrogen dioxide the beneficial effects of increased conversions based on nitrogen dioxide would be overbalanced.

If on the other hand

the yields based on propane were not materially affected or were improved then the use of oxygen would be desirable and the rising cost of propane would be of no concern. APPARATUS The apparatus consisted of a metering system for the reactants, a reactor, and a condensing system.

The flow

control for the propane and the oxygen was provided by jet type flow meters operated with the ratio of upstream to downstream pressure above the critical pressure ratio• Thus

iii. the flow was independent of small pressure fluctuations in the reactor.

Nitrogen dioxide was metered through a con­

ventional orifice meter because of the low cylinder pressure available » The reactor used was made of 14 mm. O.D,pyrex tubing wound on mandrels of 2> and 10 inches in diameter.

Two coils,

one of each diameter, were fitted together concentrically and sealed together at the bottom.

The reactor has a volume

of 2 liters. The heating of the reactor was accomplished by means of a molten salt (0 0 ^ and NaNO^) bath provided with immersion heaters operated by an automatic temperature controller. The exit gases were cooled by passage through a water condenser to remove the majority of liquid products and then through a dry ice condenser and trap to remove normally liquid products and condensable gases. Any gases remaining uncondensed were passed through a wet test meter to measure their volume and were collected over water or vented as desired. PRODUCT ANALYSIS The total liquid product was extracted with ether to separate the aqueous and non-aqueous materials.

Any

aldehydes present in each layer were determined by the hydroxylamine hydrochloride method (5 ).

The ether layer

was treated with sodium bicarbonate to remove acidic materials. The nitro paraffins remaining in the ether layer were

iv. determined "by a nitrogen analysis on an aliquot portion* The analysis used is a modification of a method described by Somers (4-)* changed*

The apparatus used by Somers was not

The reduction of the nitro paraffins was the same

as in Somer's method*

After reducing the nitro paraffins

to amines 30 ml* of distilled water was added and the flask was placed on the apparatus. NaOH was added.

Then 15 ml. of 3°% aqueous

The flask was heated by a micro burner

placed about three inches below it for 15 minutes.

During

the heating period the amines were distilled into 50 ml. of 2% boric acid solution.

They were then titrated with

0.01 N sulfuric acid using a brom cresol green-methyl red mixed indicator.

The end point is the first extremely

faint pink color seen over a white surface* Carbon dioxide, propylene, ethylene and carbon monoxide were determined by an Orsat analysis on a sample of the exit gases collected between the water and dry ice condensers*

Conversions were based on the fraction of the

nitrogen dioxide which appeared as nitro paraffins.

Yields

were based on the propane which reacted. NITRATIOH WITHOUT OXYGEH Previous work on nitrations with nitrogen dioxide did not include determinations of the by-products formed in the nitration reaction.

This fact coupled with the lack of

information about nitrations with nitrogen dioxide at temperatures between 248 and 500 C. made a series of runs without oxygen necessary.

Ho runs were made at temperatures

V

in excels of 325°0 . because indications were obtained that higher temperatures would not be beneficial• The data for these runs are given in Table I• They show that even with­ out oxygen much higher yields are possible with nitrogen dioxide than with nitric acid as the nitrating agent using present techniques. UITRATION OF PROPANE WITH-OXYGEN AT TWO MINUTES CONTACT TIME A series of runs at different temperatures was made at a contact time of two minutes and with one half mole of oxygen per mole of nitrogen dioxide.

A similar series of

runs were made using a full mole of oxygen per mole of nitrogen dioxide.

This information together with the data obtained

without oxygen shows the variations which take place with changes in temperature and oxygen feed rate when a contact time of two minutes is used.

It is to be noted that the use

of oxygen lowers the optimum temperature, improves the conversion, and increases the yield.

This can be seen in

the data presented in Tables II and III. Note that when the conversion shows a maximum the yield is also high and at or near its maximum. NITRATION OF PROPANE AT FOUR MINUTES CONTACT TIME Experiments were carried out which would furnish for a contact time of four minutes information of the same type as was found for a contact time of two minutes. This information is summarized in Tables IV and V.

It was hoped

Table I Nitrations Without Oxygen Effect of Temperature A-7*

243-2

300-2

325-1

Temperature, °C*

2k8

243

300

325

Contact time, min.

1 .2 6

1 .9 0

1.93

1.93

Prppane/NOg

4.00

3.35

4.17

4.20

Run No*

% Conversion^

13.4 —

fc Y i e l d ^

13-7

1 6 .1

l 6 .6

4?

49

51

*Dor sky ^ data ♦♦Per cent of nitrogen in nitrogen dioxide charged appearing in nitro paraffins in effluent• ♦♦♦Yield based upon moles of carbon*

Table II Nitrations with 0*5 mole Oxygen per mole of Nitrogen Dioxide Run No.

255-1

275-2

300-3

325-2

Temperature, °C.

255

275

300

325

Contact time, min#

l#93

2 .0 6

2 .0 1

1.95

Propane/NOg

4.23

4.36

4.17

4.27

ft Conversion (N)

12.9

13.3

22.9

19.9

fo Yield (Based on Propane)

60

57

53

43

vii • Table XII Nitrations with 1.0 mole of Oxygen per Mole of Nitrogen Dioxide 255-4

275-1

300-5

325-3

Temperature, °0.

255

275

300

325

Contact time, min.

1.29

2.05

1.92

2.07

Propane/NOg

4.4o

4.54

. No.

4.12

4*37

$ Conversion (N)

13.4

25.5

16.4

13.6

$ Yield (Propane)

37

62

50

4o

Table IV Nitrations with 0.5 mole of Oxygen per Mole Nitrogen Dioxide 245-1

255-2

275-3

300-6

Temperature, °C.

245

255

275

300

Contact time, min.

3 »854

4.39

4 .17

3.93

Propane/NOg

3.70

4.30

3 .68

3.62

Run No»

$ Conversion (N)

16.9

21.5

18.2

16.5

$ Yield (Propane)

66

60

60

53

viii. Table V Nitrations with 1.0 mole of Oxygen per Mole of Nitrogen Dioxide 255-3

275-4

275-9

275-7

Temperature, °C.

255

275

275

275

Contact time, min.

3.76

4 .0 3

3.6 0

3.28

Propane/N02

4.30

4 .5 0

5.20

7.30

Run No.

$ Conversion

22.6

3 2 .6

29

2 7 .5

# Yield

49

44 *

63

65

♦Results from a poor material balance so that moles of off gas have been calculated from weights charged instead of weights recovered.

that information at two contact times would show a trend towards an optimum contact time.

While the information does

seem to Toe favorable to the longer contact time the differ­ ence between the results at the two contact times is not pronounced.

Several experiments were carried out at a

shorter contact time to check this trend in favor of long contact times* NITRATION OF PROPANE AT ONE AND ONE HALF MINUTES CONTACT TIME The runs made at one and one half minutes contact time were all made with a full mole of oxygen per mole of nitrogen dioxide.

The information from these runs again

shows very little advantage of one contact time over another, but the longer contact times are somewhat favored. effect is more pronounced in the yield figures.

This

The data

for these runs is presented together with the data from the runs at the other two contact times in Tables VI and VII. The tables show the effects of varying the contact time, it can be seen that the optimum temperature is lower at the longer contact time. NITRATION OF PROPANE AT THREE MINUTES CONTACT TIME The information already obtained was used to pick the conditions for a single run at a contact time of three minutes, 235°C., and three quarters of a mole of oxygen per mole of nitrogen dioxide.

This run resulted in a con­

version of 29 $ and a yield of 71%, the highest yield and nearly the highest conversion obtained in this investigation.

X.

Table VI Variations in Conversion with Temperature and Contact Time Og/EOg

-

1 Conversions at Times Shown 2 min* 4 min 1*5 min*

Temperature, °C* 255

-

13 A

22.6

275

1S.4

25*5

29*0

300

25.4

16*4

22*5

325

1 7 .1

1 3 .6

—

Table VII Variations in Yield with Temperature and Contact Time 02/TO2 Temperature, °C*

=

X Yields at Contact Times Shown 1*5 min* 2 min* ^ min*

255

-

37

^9

275

55

62

6?

50

4g

300

325

35

BY-PRODUCTS High yields of nitro paraffins reduce the importance of the nature of the by-products in nitrogen dioxide nitrations as compared with nitric acid nitrations using present techniques* Carbon dioxide and carbon monoxide constitute the main by­ products with carbon dioxide predominating.

The yields of

propylene, ethylene, and carbonyl compounds are of minor significance under good operating conditions and usually total less than 10$ based on the propane when optimum yields of nitro paraffins are obtained* ATTEMPTS TO CATALYZE THE REACTION Usually attempts to catalyze the nitration reaction increase the formation of oxidized products.

It was thought

that at the low temperatures employed in nitrations with nitrogen dioxide this generalization might not hold, but the use of ferric oxide and boric oxide resulted in lower yields and conversions to nitro paraffins* The addition of small amounts of alcohols to the feed in the nitration reaction was also unsuccessful in that lower yields and conversion to nitro paraffins were obtained* NITROGEN BALANCE Addison (l) discovered that in nitrations with nitric acid about twenty per cent of the nitrogen in the nitrating agent was lost for recycling in the form of nitrogen gas.

As a result of this discovery it was decided

to determine free nitrogen in the exit gases from the nitration of propane with nitrogen dioxide.

The analyses

xii. showed that about three per cent of the nitrogen charged in the form of nitrogen dioxide was lost in the form of nitrogen gas• Further attempts were made on runs previously completed to show by nitrogen balance that high yields of nitro paraffins could be obtained based on the nitrating agent♦ This was not successful because no distinction had been made between nitrous acid and carboxylic acids which are both formed in small amounts in the nitration reaction. It can be stated, however, that the formation of free nitrogen is small, and that this is a definite advantage of nitrogen dioxide over nitric acid as a nitrating agent. MECHANISM The results of this research do not conflict in any way with the mechanism proposed by Addison (l) for the nitration of butane with nitric acid.

Briefly this theory

is that nitroparaffins are formed by the reaction of hydro­ carbon free radicals and nitrogen dioxide to form a stable nitro paraffin or an alkyl nitrite which is instable at the temperatures of the nitration process. The decomposition of alkyl nitrites leads to the formation of aldehydes, nitric oxide, and lower hydrocarbon free radicals, which in turn may be nitrated.

The formation of carbon monoxide results

from the decomposition of aldehydes.

The long contact times

used in this research lead to decomposition of most of the aldehydes•

xiii. ACKNOWLEDGMENT The authors are indebted to the Commercial Solvents Corporation and the Purdue Research Foundation for financial assistance in carrying out this research.

xiv LITERATURE CITED (1)

Addison, L. M., A Study of the Vapor Phase Nitration of Butane with Emphasis on the Effect of Adding Oxygen, Ph. D. Thesis, Purdue University (1950)*

(2) Alexander, L. Gr., A Study of the Vapor Phase Nitration of Paraffins, Ph. D. Thesis, Purdue University (19^7). (3) Dorsky, J., Vapor Phase Nitration of Propane with Nitrogen Dioxide, Ph. D. Thesis, Purdue University

(1940).

(4)

Somers, P. D., Determination of Nonaainoid Nitrogen in Aliphatic and Aromatic Compounds, Proc. Indiana Acad., Soi., 117-20 (19^5)•

(5)

Walker, J . F., *Formaldehyde", p. 263 , Reinhold Publishing Co., New York, New York (1944).

TABLE OF CONTENTS

Page ABSTRACT (Nitrations with Nitric Acid)..........

i

ABSTRACT (Nitrations with Nitrogen Dioxide)......

i

PART I NITRATIONS WITH NITRIC ACID INTRODUCTION..................................

1

DISCUSSION OF RESULTS..........................

2

Nitrations with Oxygen and High Surface......

2

Nitrations with Steam, High Surface and Oxygen.

5

Effects of Low Propane to Oxygen Mole Ratios...

2>

The Nitration of Propylene .................

12

Effect of Adding Bromine in Nitration........

l4

EXPERIMENTAL..................................

27

Apparatus and Technique.......... SUMMARY.......................................

2J 23

PART II NITRATIONS WITH NITROGEN DIOXIDE INTRODUCTION..................................

JO

DISCUSSION OF RESULTS..........................

J2

Nitrations without Oxygen...................

34-

Nitration at Two Minutes Contact Time with Oxygen......

37

Nitrations at Four Minutes Contact Time.......

4-2

Nitrations at One and One-Half Minutes Contact Time....

4-6

Nitration at Three Minutes Contact Time.......

4-9

By-Products................................

50

Yield Based Upon Nitrogen Dioxide............

92

Page Mechanism.. ...............................

5^

Attempts to Gtalyze the Nitration Reaction..... EXPERIMENTAL..................................

6l

Apparatus and Technique.....................

6l

Product Analysis

63

....................

Calculations...............................

79

Chemicals Used..............................

31

SUMMARY.......................................

32

BIBLIOGRAPHY..................................

33

VITA........... ..............................

LIST OF TABLES PART I NITRATIONS WITH NITRIC ACID Table

Page

I

Effect

of Varied Oxygen Feed Rates.............

3

II

Effect

of Varied Nitric Acid Fead Rates..........

6

III

Effect

of Increased Oxygen Feed................

9

Nitration of Propylene........................

13

The Effect of Bromine on The Nitration Reaction Without Oxygen..........

16

Effect of Using Bromine with Oxygen in the Nitration of Propane with Nitric Acid..........

IS

The Effect of Varying Oxygen Feed in Nitrations with Bromine......

20

IV V VI VII

PART II NITRATIONS WITH NITROGEN DIOXIDE I

Nitrations Without Oxygen Effect of Temperature.•

35

II

Nitrations with 0«5 mole Oxygen per mole of Nitrogen Dioxide.............................

3^

Nitrations with 1.0 mole of Oxygen per mole of Nitrogen Dioxide..............................

40

Nitrations with 0*5 mole of Oxygen per mole of Nitrogen Dioxide.............................

43

Nitrations with 1.0 mole of Oxygen per mole of Nitrogen Dioxide..............................

45

III IV V

IVa Variations in Conversion with Temperature and Contact Time.......... Va Variations in Yield with Temperature and Contact Time................................ VI VII

4% 4%

Iron Oxide Catalyst Compared to non-Catalytic Operation............

57

Boric Oxide Catalyst Compared to non-Catalytic Operation ..............................

59

VIII Effects of Adding Alcohol to The Feed Without Oxygen............

60

Table IX

page Effects of Adding an Alcohol to The Feed With Oxygen ...................................

60

Summary of Primary Data for Nitrations with Nitrogen Dioxide.............. ...............

LIST OF FIGURES PART I NITRATIONS WITH NITRIC ACID Figure 1 2

page Effect of Varying Oxygen Feed in Nitration with Bromine.................................

20a

Variation in Moles of Product with Oxygen Feed with Bromine..........................

20b

IMPROVED PROCEDURES FOR THE VAPOR PHASE NITRATION OF PROPANE

Part I Nitrations with Nitric Acid

INTRODUCTION The beneficial effect of oxygen upon the conversion of nitric acid in the nitric acid nitration of propane was noted by Alexander (2).

Addison (l) investigated the effect

of oxygen with high reactor surface using butane. investigated the effect of oxygen with steam.

He also

The present

investigation was planned to study the effect of the similtaneous use of oxygen, steam, and high reactor surface on the nitration of propane.

It was also thought important to

secure data on yields based upon propane.

This aspect of

the work was completely ignored by Alexander but it has become important as a result of the increase in cost of propane. As a complete review of the literature which serves as background to this problem is to be found in the theses of Alexander and Addison it will not be repeated here. Throughout this investigatorfs work under Dr. Bachman, he has persistently suggested that use be made of additional methods of creating free radicals besides the addition of oxygen.

This was suggested in order to avoid

some of the oxidation processes which result from the use of oxygen and which lead to numerous by-products.

He

suggested the use of bromine for this purpose and the effect of bromine has been investigated.

2*

DISCUSSION OF RESULTS I»

Nitrations with Oxygen and High Surface The apparatus used was the same apparatus used by

Addison (l)#

In some cases the size and calibration of

metering jets had to be adjusted, but the preheater and reactor were exactly the same. 28/1 S/V reactor.

The reactor used was Addison's

This was selected because Addison had

made two runs with propane which were the same except for a change in surface to volume ratio and which showed that better results would be expected at 28/1 S/V than at 300/1 S/V.

Since the aim of this research was to develop suitable

commercial methods the initial run was made with a small amount of oxygen since it was to be expected that the con­ version without oxygen would be low.

The selection of a

temperature of 425°0 was based mainly on the fact that the reactor being used had been shown to have poor external heat transfer characteristics and this temperature would be necessary to achieve about the same results as Alexander found at 4lO°C with a reactor having excellent heat transfer characteristics. The contact time was chosen to coincide with that of Alexander (2) and Addison.

In order to find

the effect of adding steam it was first necessary to study the reaction without steam.

This was accomplished with

the results summarized in Table I. observed by Alexander was found.

The same general effect The addition of oxygen

3Table I Effect of Varied Oxygen Feed Rates 1

2

4

Temperature, °C.

^25

425

425

Contact time,, sec.

1.5

1.5

1.7

Rim

28

28

< y °2

21.6

10.0

cyHHOj

11.0

12.5

S/V Mole Ratios

28 7.2 20.

Og/mro

0 .5 1

1.2 5

2 .8(

HgO/HNO,

1.42

1.42

1.4:

Conversion % (N)

4l

51

58

Yields #

39-2

34.0

3 4 .3

4.1

2.9

1.7

2 1 .7

3 0 .8

33.9

Ethylene

6.9

7.0

2 .5

CO

9.3

7.3

9.7

18.9

18.0

17.7

RHOg CV1

8

Propylene

Carbonyl Compounds

increased the conversion and the more oxygen added the higher was the conversion.

In neither Alexander *s nor the present

work was a condition observed in which a large amount of oxygen would cause the conversion to decrease after reaching a peak.

This does not prove that such a phenomenon does

not exist, in fact, it appears from Alexander1s data that he was near the peak.

The runs summarized in Table I show

that in no case was a high yield based on propane obtained. The formation of each by-product increased with the increase in conversion to nitro paraffins.

The conversion to carbon

dioxide, ethylene and oarbonyl compounds did not increase as fast as the conversion to nitro paraffins.

The conversion

to propylene increased much faster than that to nitro paraffins♦ The yields of these products are shifted accord­ ingly.

A sufficient background having been created attention

was next turned to the use of steam diluent*

5 II.

*

Nitrations with Steam, High Surface and Oxygen It has been seen that relatively high mole ratios

of oxygen to nitric acid and attendant low mole ratios of propane to oxygen are desirable to effect high conversions in the nitric acid nitration of propane• Of these two ratios it seems more profitable to think in terms of the ratio of propane to oxygen for the evidence is strong that the role of oxygen is primarily concerned with the low temperature oxidation of propane to produce hydrocarbon free radicals. The use of steam was tried because Addison's work on butane indicates that such would improve the yield based upon propane. Since the effect of steam upon the balance of temperature and contact time was not known it was decided to make a series of runs in which the temperature and the ratio of propane to oxygen were held constant and the quantity of nitric acid charged was varied.

This also held the

ratio of steam to nitric acid constant, and, incidentally, varied the ratio of hydrocarbon to nitric acid.

Table II

seems a little irregular as far as the distribution of by-products is concerned.

This results from the substitution

of run 9 in the place where run 7 was intended.

Run 9 differs

from run 5 and 6 in that the preheater temperature had to be raised from 265° to 350 °^ in order to bring about more steady operation.

This unsteady operation was noted to a

6 Table II Effect of Varied Nitric Acid Feed Ratea g

5

9

6

Temperature, °C•(Preheat)

3^0

265

350

265

temperature. °C.(Reactor)

425

iJ-25

425

425

Contact time,, sec*

1*3

1.4

1 .8

1*9

Run

28

S/V Mole Ratios

28

o3 /o2

5*1

5.1

5"1

Cj /HNO

8 .6

9.7

1 6 .8

Og/HNO-

1*7

1.9

3*3

HgO/HNO,

28 5-1 20

3*9

15

15

15

15

35

54

62

54

BHOg

31*7

37.3

37.6

36 .6

00g

0 .6

1-3

0 .8

2.6

2 8 .7

2 7 .0

26.6

1 5 ,3

Ethylene

1 .2

3*2

0.8

1 1 .8

CO

7.3

8 .7

9.3

6*9

30 .5

21.9

25 .1

2 6 .2

Conversion fo (H) Yields ft

28

Propylene

Carbonyl Compounds

7 lesser extent in runs 5 and 6 and was traced to the inability of the preheater to do its job when called upon to vaporize the water injected into it through the nitric acid jet.

The

appearance of a peak conversion under the conditions of run 9 is to be expected from, the conversions for runs 5 and 6. In all the runs with steam less olefins and more aldehydes are formed than in similar runs without steam. At optimum conversion the conversion is slightly above that expected without steam, and the yield on propane is signi­ ficantly higher.

The use of steam does not affect the

relationship between contact time and temperature from that without steam.

III.

Effects of Low Propane to Oxygen Mole Ratios

It was hoped that the use of steam would allow the advantageous use of more oxygen than was used in the series of runs at a propane/oxygen mole ratio of ^.1.

To test this

idea a series of runs was made in which the propane/oxygen mole ratio was changed from ^.l to 4-.0 and then to ^.2. This is near the limit of combustibility with no diluent present.

The apparatus was modified so that the oxygen

could be introduced into the stream after the steam diluent had been introduced.

This would have allowed the use of

even greater amounts of oxygen, but the results of the three runs showed that further increases in oxygen content of the feed would be undesirable. It will be noted in Table III that the change in propane/oxygen ratio is accompanied by an increase in the oxygen/nitric acid ratio.

The results of these changes is

a decrease in both the yield and conversion.

There is no

apparent reason to expect a reversal of this trend if still more oxygen were to be used.

If we consider the effects of

the use of more oxygen on the by-products we note a number of things. The yield of carbon dioxide rises.

The formation

of carbon monoxide is also favored by more oxygen.

The

trend seems a little irregular because of what appears to be a slightly low figure for run 10.

Total olefin formation

decreases as the oxygen is increased, but this desirable

9 Table III Effect of Increased Oxygen Feed 9

10

11

Temperature, °0.

425

^25

425

Contact time,, sec#

1.8

1.8

1*9

Run

28

8/7 Mole Ratios

28

28

c3/'°2 C3/HHO,

5-1

4.0

3*2

16*8

14.1

1 2 .1

o 2/h b o 3

3-3

3*5

3*9

HgO/HHO,

1 5 .O

1 5 .O

1 5 .O

Conversion fo (H)

62

57

52

Yields #

BSOg

37*6

36.1

35*2

co2

0 .8

1 .2

2 .2

2 6 .6

2 6 .0

2 2 .0

Ethylene

0 .8

0 .6

1.2

CO

9.3

8 .1

l4.0

2 5 .I

27 .9

25 .5

Propylene

Carbonyl Compounds

*

10. effect is much more than off set by the increase in the oxides of carbon.

It seemslogical that olefin production appears

to go down because propylene is produced only to be oxidized further. increased.

Ethylene production seems to rise as oxygen is Aldehydes, which are the most valuable by-products

are shown by the data to increase slightly to a peak and then decrease again. In every run a good oxygen balance can be obtained assuming that each olefin formed produces one mole of water and thus uses one-half mole of oxygen.

As more oxygen is

added to the feed more oxygen reacts.

After enough oxygen

has been added to reach the peak conversion the additional oxygen does one of two things.

It can either form more by­

products in general, or it can shift the character of the by-products formed to those having a higher oxygen content or a lower hydrogen content.

The first of these actions

takes place and is evidenced by the fact that a smaller proportion of the total product is found as nitro paraffins. The second also takes place.

The shift to the more highly

oxidized by-products has already been noted. These results are not necessarily the same as would be expected without steam.

In operation without steam a

stricter limit of inflammability must be observed.

This may

be the reason why Alexander (2), who approached this situation in operation without steam, did not continue in this direction. It may also be that Alexander felt that he was already near

11. the peak conversion to nitro paraffins.

His data show a

sharply diminishing increase as more oxygen is added• After passing the peak conversion the situation with regard to the by-products is the same as with steam.

12. IV.

The Nitration of Propylene

The vapor-phase nitration of propylene is not recorded in the literature.

Aside from this fact it is of

interest because of its probable occurence in the commercial nitration process.

Propylene is formed as a by-product and

is recycled along with the unused propane.

The recycling

of propylene is practically unavoidable because its separation from propane is difficult and costly. Only one run was made on the nitration of propylene. The results of this run are given in Table IV.

Under the

column headed 9(12) the yields of various products are listed on the assumption that the propylene in run 9 is recycled. It will be noted that this experiment was carried out under the same conditions as run 9*

13 Table IV Nitration of Propylene Run

12

Temperature, °C*

425

Contact time, sec*

1*8

S/V Mole Ratios

28 Propene/Og Propene/HNO^

5*1 18*2

Og/HNCy BgO/HNO^ Conversion Yields fa

9(12)*

3*6 15*0

(N)

25*8

RN02

9*8

40.2

co2

0.6

1.0

Ethylene

37 *3

10*6

CO

28.0

16.8

Carbonyl Compounds

24.2

31.5

♦The yields are listed which would result if the propylene formed in run 9 were recycled.

These values are calculated

with the aid of data from run 12*

lA. 7.

Effect of Adding Bromine in Nitration The use of bromine in the nitration process was

suggested by Dr* Bachman*

He pointed out that bromine might

be expected to dissociate into bromine atoms and to react in this form with hydrocarbons to produce free radicals even more readily than would oxygen or nitric acid*

Bromine would

offer the further advantage of being regenerated continuously (as bromine atoms) by the oxidizing agents present in the nitration process*

This would serve to maintain a high rate

of formation of free radicals using only a small amount of bromine in the feed*

Furthermore, any organic bromides

formed would decompose in the reactor at the temperatures used*

This would form hydrogen bromide and olefins*

The

hydrogen bromide would in turn be oxidized to bromine atoms* A similar regeneration of active particles cannot occur when oxygen is used to produce the free hydrocarbon radicals since any oxygen containing compounds such as water, alcohols, aldehydes, and acids are not oxidized by nitric acid to free oxygen*

The details of this process will be considered at

the end of this section* A run with steam and high surface but without oxygen was made for comparison purposes*

This allowed the effects of

bromine to be observed independent of the effect of oxygen* Steam was introduced by using dilute nitric acid.

The dilute

acid also served as a solvent for the bromine used in the

15. comparative run made later.

The results of the two runs with

and without bromine are shown in Table V.

lés Table V

The Effect of Bromine on the Nitration Reaction Without Oxygen Run

14-

15

Temperature, °Q#

423

423

Contact time, sec*

1.3

1 .9

S/V Mole Ratios

C-/HNO, j y HgO/HHO, Brg/HNO-

Convereion %

20.4

22.3

15 *0

1 5 .0

0

0.015

27.7

27.4

5 0 .0

2-7

none

Propylene

24*3

4*5

Ethylene

23*4

15.0

CO

12.6

19.5

Carbonyl Compounds

9 .5

11.0

RN02

0.0425

0.0459

co2

0.0119

none

Propylene

0.0353

0.0039

Ethylene

0.0517

0.0196

CO

0.0556

0.0510

Carbonyl Compounds

0.0349

0.0240

bho 2

co2

Moles Recovered

23

23.0

(H)

Yields on Propane

23

17. It will be noted that the use of bromine improved the conversion somewhat• The more spectacular effect was upon the yield and the distribution of by-products*

The

yield of 50 $ is by far the highest found up to this point in the investigation* 39$ found in run 1*

The best previously reported figure is the Although yield figures are given in the

table the results of the run are so different that it is more informative to examine the actual moles of products produced. No carbon dioxide was found with bromine present * Propylene production was cut to less than one-ninth by the use of bromine*

The reduction in the amount of ethylene was the

next most striking feature with less than half as much being found.

Carbonyl compounds were reduced by a third and a

small reduction was found in the quantity of carbon monoxide formed* While the use of bromine without oxygen vastly improved the yield based upon propane its commercial possi­ bilities would be considerably enhanced if higher conversions could also be obtained.

Thus it was decided to use bromine

under the same conditions which gave the best results without bromine.

This situation is found in run 9*

bromine under identical conditions is run 16. runs are compared in Table VI.

The run with Those two

Table VI Effect of Using: Bromine with Oxygen in the Nitration of Propane with Nitric Acid 9

16

Temperature, °0»

423

4-23

Contact time, sec*

l.g

1.8

Run

23

S/V Mole Ratios

Conversion $

28

5*1

5.1

O-j/HHOj

1 6 .3

1 6 .5

o2/hno.

3*3

3.1

HgO/HKO,

1 5 .0

1 5 .0

Brg/HNO,

0

0-/0,

2

(H)

Yields on Propane

6 1 .7

4-6.0

37 •6

33.2

0 .8

0.5

2 6 .6

1 6 .I

Ethylene

0 .8

7-6

CO

9.3

13.3

25.1

29.4.

RNOg co2 Propylene

Carbonyl Compounds Moles Recovered

0.0 15

rno2

O.I378

0.1 03 0

co2

0.0081

0.0042

Propylene

0 .0928

0.0473

Ethylene

0.0040

0.0334

CO

0 .0 9 6 8

0.1 17 0

Carbonyl Compounds

0.2 19 0

0 .2 1 7 0

19 • Quite obviously these conditions are too drastic for the use of bromine, but it is worthy of note that bromine reduced the formation of propylene and carbon dioxide and held carbonyl compounds to the comparative level.

Having estab­

lished that too much oxygen was used in run 1 6 , runs 1J and 13 were made with intermediate amounts of oxygen.

These

runs are reported in Table VII which gives the conditions and results for all the runs with bromine. Note that as the oxygen is decreased from that required for a propane oxygen ratio of 5*1 to 3.2 that both yield and conversion rise. Although the yield of 57% found in run 13 is the highest reported, run 17 with a conversion of 4-7.7% and a yield of 55*5 Is the best balance of yield and conversion. ately the contact time in run 17 was low.

Unfortun­

This resulted

from a high flow of nitric acid following the removal of a plug of Silicon grease encountered when the jet was removed from the apparatus at the close of run 1 6 . Note that in runs of equal length in the same apparatus, run 17 produced O .1765 mole of nitro paraffins while run 9 with a higher conversion produced only 0 .137 ^ mole of nitro paraffins. The formation of by-products show several constant trends.

Propylene formation increases with oxygen concen­

tration but is comparatively low throughout♦ Carbon mon­ oxide formation also increases with oxygen after a remarkable drop with the first addition of oxygen.

The analysis of the

off gas for carbon monoxide in run 13 was checked twice.

20. Table VII

The Effect of Varying Oxygen Feed in Hitrations with Bromine 15

18

17

16

Temperature, °0.

423

4-23

4-23

423

Contact time, sec.

1 .9

1-7

1 .5

1 .8

Run

23

S/V Mole Ratios

28

28

28

-

11 .5

8 .2

5.1

C3/HNO2

2 2 .8

1 0 .6

9*9

1 6 .5

Og/HHO^

0

1 .2

3.1

1 5 .0

1 5 .0

C^/Og

HgO/HHO^

1 5 .0 0.0 1 5

Brg/HHO^

O .92 15*0 0.015

0.0 1 5

0 .0 1 5

Conversion ^ (n)

27.7

4.3 .2

4-7*7

46.0

Yields on Propane RNO^

5 0 .0

57-0

55*5

3 3 .2

co2

none

0 .2

none

0 .5

Propylene

4-.5

7-3

9*7

1 6 .1

Ethylene

1 5 .0

7 .8

4-.1

7.6

CO

19.5

1 .0

3 .6

1 3 .3

Carbonyl Compounds

11 .0

2 6 .6

2 7 .O

2 9 .4

Moles Recovered^ RliOg COg

0.04-59 none

0.1371

O .1765

0 .1 0 3 0

0.0017

none

0.0042

Propylene

O.OO39

0.0164-

0.0294

0.0473

Ethylene

0.0 1 9 6

0.0262

0.0184

0.0334

CO

0 .0 5 1 0

O.OO65

O.O33 I

0 .1 1 7 0

Carbonyl Compounds

0.0240

0.14-90

0.2445

0 .2 1 7 0

♦All adjusted to same run time.

Figure 1

EFFECT OF V A R Y IN G O X Y G E N F t E u IN N I T R A T I O N W IT H B R O M I N E 00oC. Later he found that lower temperatures were more desirable and he completed a series of runs at 24g°C with varying contact times.

He did not measure the formation of by­

products formed from propane.

His best result was a con­

version of 26% achieved at 24-g°C and l4 minutes contact time.

The broad range of temperatures from 24g°C to

500°C was not investigated. Since Dorsky*s work several patents have been issued to Levy (6) in which he claims a catalytic effect in the nitration reaction.

The catalysts suggested

included oxides of arsenic and antimony, glasses containing these oxides, and sodium arsenite.

These catalytic effects

have never been confirmed although numerous attempts were made in this laboratory by Shecter (8). Alexander (2) made two runs using propane, oxygen, and nitrogen dioxide.

These runs were of a preliminary

nature using high temperatures and short contact times.

His

31. results with nitrogen dioxide were such as to indicate that under the proper conditions nitrogen dioxide might be a good nitrating agent. This research is an outgrowth of Alexander's preliminary runs.

32 Discussion of Results The selection of the type of apparatus and the operating conditions to be used were governed in large part by the desire to develop any commercial possibilities*

Thus,

although an analogy with nitric acid nitration would indicate the advantage of a higher mole ratio of propane to nitrogen dioxide an arbitrary ratio of 4 to 1 was chosen*

This ratio

conformed to that used by Dor sky (4) and allowed comparison with his work at 243°0•

This ratio was chosen also because

it would allow the study of reasonably long contact times and still produce a large enough sample of product for analysis in a reasonable period of time using a reactor as large as the available salt bath would accomodate* In order to construct a reactor of sufficiently large volume it was obvious that larger diameter tubing than that used in other nitration reactors would have to be used* It was found that 14 mm. O.D* tubing was about as large as could be wound into a coil conveniently*

Two concentric coils

of this tubing connected together made up the reactors* The accurate metering of oxygen and propane was accomplished with the aid of jets operated so as to permit terminal velocity in their throats*

The operation and

principle of this method are described under Apparatus. All conversions were determined on the basis of modified Kjeldahl-method of analysis for nitrogen.

Hydro­

carbon yield calculations were made from analyses for the

33 principal hydrocarbon by-products, namely, aldehydes and acetone, olefins and oxides of carbon.

Analyses of the

products also determined oxides of nitrogen and acids. analytical procedures are described under Experimental Procedure.

The

34 le

Nitrations Without Oxygen

At the outset of this work it seemed desirable to be able to compare the work with that of Dorsky (4)•

The

first run was made to duplicate the condition of one of Dorsky18 runs.

This allowed acheck of

his work and yet was

so selected as to be a part ofa series of runs in which the temperature was to be varied* Dorsky did not analyse each run but measured the volume of products from each run.

He combined all the

products of a series of runs, and assigned a fraction of the total nitro paraffins found to each run proportionately. Originally, distillation of the nitro paraffins from the water layer was used to separate them from the water solutions This proved too inefficient on a small scale but accounts for a low conversion in the first attempt to duplicate one of Dorsky*s runs. The results of nitrations at 2 minutes contact time without oxygen are given in Table I.

A contact time of 2

minutes was used at first because it was felt that this represented a significant shortening of contact time from the 14 minutes necessary for Dorsky*s best conversion and about as short a contact time as could be expected to produce high conversions.

35 • Table I Nitrations Without Oxygen Effect of Temperature Run No.

A-7^

248-2

300-2

323-1

Temperature, °0.

243

248

300

325

Contact time, min.

1 .8 6

1.90

1.93

1-93

Propane/NOg

4.00

3 .8 5

4.17

4.20

$ Conversion^ Yield.***

13.4 -

I?.?

1 6 .1

1 6 .6

4?

49

51

♦Dorsky*8 data ♦♦Per cent of nitrogen in nitrogen dioxide charged appearing in nitro paraffins in effluent• ♦♦♦Yield based upon moles of carbon.

Since both conversion and yield seemed to be still rising with temperature it might be asked why still higher temperatures were not investigated• It was suspected that the use of oxygen would shift the optimum temperatures at a given contact time. temperature.

The most logical shift was to a lower

Subsequent work proved this to be true.

Recourse to higher temperatures was therefore unnecessary. Since the conversion is still quite low at 325° any further increase in temperature does not seem promising.

As the

temperature was raised the rate of increase in the conversion decreased indicating that the maximum is not far from 325°0 . The yield tends to follow the conversion.

This

effect has been observed to a great extent in the experiments with nitrogen dioxide at low temperature.

The effect is

primarily due to the fact that the conditions for nitration are more critical than for by-product formation.

Thus, if

conditions are changed to produce more nitro paraffins, a higher per cent of the hydrocarbon is converted to nitro paraffins.

It is the relative conversion to the various

products which determine the yields of those products. Thus it is seen that without oxygen it would not be worth while to study temperatures above 325°0 with the technique used in this experiment. Adequate comparative data was at hand for comparison with runs containing oxygen.

37 II.

Hitration at Two Minutes Contact Time With Oxygen After the work without oxygen at 2 minutes contact

time it was felt to he desirable to retain this contact time and study the same temperature with two different oxygen contents of feed.

These experiments confirmed the idea that

oxygen content varied the temperature at which optimum conversion would he obtained.

This complicated the work to

the point that a great deal of time was spent studying the relationships of contact time, temperature and oxygen/nitrogen dioxide mole ratio.

For not only did optimum temperature

at one contact time vary with the addition of oxygen, but there existed the possibility that a decided optimum match of contact time and temperature would appear when oxygen was present• It should be noted that as the mole ratio of oxygen to nitrogen dioxide is changed and the hydrocarbon to nitrogen dioxide ratio is held constant, the hydrocarbon to oxygen mole ratio is also changed.

The range of ratios

studied are about the same as have been studied in the case of nitrations with nitric acid. In Table II the results are given for a series of runs using 0.5 mole of oxygen at various temperatures. the optimum temperature is at or near } ’00°0»

Here

The yield holds

nearly constant as the temperature is increased but falls off sharply above the temperature of maximum conversion.

38

Table II Nitrations with 0.5 mole Oxygen per mole of Nitrogen Dioxide 255 -I

275-2

3 OO-3

325-2

Temperature, °0.

255

275

300

323

Contact time, min.

1.98

2.0 6

2.0 1

1.95

Propane/NOg

4.23

4 .5 6

4.17

4.27

Run No*

io Conversion (N)

12.9

13.3

23.9

1 9 .9

$ Yield (Based on Propane)

60

57

58

43

39 The best operating temperature for this contact time and 0»5 mole of oxygen per mole of nitrogen dioxide brought about a large increase in conversion and yield.

This

conversion obtained was over 3% than Dorsky*s best conversion in one-seventh of the contact time.

This was obtained in

combination with a yield which is 7$ better than the best obtained without oxygen.

These results for 0 .5 mole of

oxygen were so favorable that it was decided to try a full mole of oxygen.

The results of this series of experiments

are given in Table III.

lote that increasing the amounts

of oxygen used lowers the optimum temperature.

It also

appears that optimum conditions are more critical using more oxygen since the drop in yield and conversion per degree is greater the more the oxygen content of the feed.

(Compare

1.0 mole with 0*5 mole and with no oxygen). The actual optimum temperature is near 275°C and probably slightly above this temperature.

ko Table III Nitrations with 1.0 mole of Oxygen per Mole of Nitrogen Dioxide 255-4

275-1

300-5

325-3

Temperature, °0

255

275

300

325

Contact time, min.

1.8 9

2.05

1 .9 2

2.07

Fropane/NOg

4.4o

4.54

4.18

4.37

i No.

$ Conversion (N)

13.4

25 .5

1 6 .4

1 3 .6

“ fa Yield (Propane)

37

62

50

4o

In general, however, this series of runs was disappointing for very little improvement was realized for the extra oxygen.

Hext it was hoped to secure higher

conversion at lower temperature by using a longer contact time.

42.

III.

Nitrations at Four Minutes Contact Time

Since Dorsky1s work without oxygen showed only a moderate increase in conversion with increase in contact time, it was felt that little short of doubling the previously used contact time would produce significant change*

It was

hoped that the nature of nitrogen dioxide would produce a higher conversion at the best temperature for a long contact time in comparison with the best temperature at a shorter contact time*

In other words there existed the possibility

of an optimum contact time. The data collected using 0*5 mole of oxygen is found in Table IV.

With one half mole of oxygen per mole of

nitrogen dioxide higher yields were obtained in general than were obtained at 2 minutes contact time, but the conversions as measured were not as high*

Most likely the conversions

at a temperature slightly above 255°0 would be comparable with the best obtained at 2 minutes contact time since the conversion drops sharply in lowering the temperature 10 degrees and drops only moderately in a 20 degree rise*

This

suggests that the curve is still rising at 255 °C and reaches a peak between 255 and 2J5°G and is decreasing at 275°0 . Even though this peak be considerably higher than the 2 1 .5% found at 255 , it fails to suggest a conversion worthy of consideration with the decrease in productivity necessitated by the long contact time*

tr­ iable IV Nitrations with 0.5 Mole of Oxygen per Mole Nitrogen Dioxide Run No.

245-1

255-2

275-3

300—6

Temperature, °C.

245

255

275

300

dontact time, min.

3.84

4.3 9

4.17

3-98

Propane/NOg

3.70

4 .3 0

3 .6 8

3 .6 2

% Conversion (N)

16.9

21 .5

18.2

1 6 .5

^ Yield (Propane)

66

60

60

53

H4. There remained the possibility that at longer contact times optimum conditions would require more oxygen.

Table V summarizes the runs made at longer contact

time and using a full mole of oxygen.

All the runs in the

table are not comparable• Runs 255-3 and 275-9 are comparable. They show that the maximum conversion is near 275°0*

Ex­

perience with less oxygen suggests that higher temperatures would not be desirable.

Run 275-^ shows a higher conversion

than Run 275-9 because of the longer contact time*

The

yield shown is low because a poor material balance was secured.

Since the weights of normally gaseous products

recovered were very low the weight of off gas used was adjusted to fit the weight of material charged.

This is

undoubtedly conservative and represents a sort of minimum value.

At best, however, a conversion of 32.6^ and a

yield of about 63$ might be achieved.

Run 275-7 demonstrated

the futility of efforts to obtain high conversions by increasing the mole ratio of propane to nitrogen dioxide for the desirable effects of this change are more than off set in comparison with Run 275-9 by the change in contact time.

In general it is concluded that the use of

a contact time of 4 minutes is not desirable from a commercial standpoint in comparison to the 2 minute contact time.

%5. Table V Nitrations with 1.0 Mole of Oxygen oer Mole of Nitrogen Dioxide 255-3

275 -iJ-

275-9

275-7

Temperature, °0.

255

275

275

275

Contact time, min.

3.76

4.03

3.60

3 .2 d

Propane/NOg

4.30

4.50

5-20

7*30

Run No*

$ Conversion

2 2 .6

3 2 .6

29

27*5

io Yield

%9

#♦

63

65

♦Results from a poor material balance so that moles of off gas have been calculated from weights charged instead of weights recovered.

46 IV*

Nitrations at One and One-half Minutes Contact Time

The reactions run at a contact time of four minutes showed that the maximum conversion obtainable was nearly the same as that which was obtained at two minutes contact time* This fact, in connection with the commercial advantages of a shorter contact time, suggested the possibility of using a contact time shorter than two minutes.

A contact time of

one and one-half minutes was chosen because this was just about as short a contact time as was permitted by the existing flow measuring devices*

It was possible to change

the capacity of the flow meters to allow even shorter contact time if a contact time of one and one-half minutes showed promise* Three runs were made in this series.

All of them

used one mole of oxygen per mole of nitrogen dioxide. runs

These

showed the usual effects of temperature in that an

optimum near 300°0 was formed. A good comparison of data of three contact times is shown in Tables IVaand TB. All of these data are from runs using one mole of oxygen per mole of nitrogen dioxide• The conversion data show that the best conversion obtainable at a given time by varying the temperature is approximately equal to that obtainable at another contact time if the temperature is shifted to the best value for the new contact time*

As would be

47Table IVa Variations in Conversion with Temperature and Contact Time

Og/NOg = 1 Temperature,

Conversions at Times Shown 1#5 min. 2 min. 4 min.

C.

255

~

15*4

22 .6

275

13.4

2 5 .5

2 9 ,0

300

2 5 .4

1 6 .4

22 .5

325

17.1

13*6

Table Va Variations in Yield with Temperature and Contact Time

Og/KOg = 1 Temperature, °C.

Yields at Contact Times Shown 1 .5 min. 2 min. 4 min.

255

-

37

49

275

55

62

63

300

49

50

48

325

35

4o

-

expected- the optimum temperature increases with a decrease in contact time.

If any trend exists which favors one

contact time over another the trend is in the direction of longer contact times.

This is undesirable from a commercial

standpoint. Two effects influence the yield data for nitrations with nitrogen dioxide.

The yield tends to rise with the

conversion since fewer passes are required to consume the hydrocarbon and thus the extent of by-products formation is decreased.

The effect of raising the temperature is to

produce more oxidation and correspondingly less nitro paraffins.

Above the optimum temperature both effects tend

to lower the yield, but below the optimum temperature they tend to counteract each other.

Most of the conditions used

resulted in a coincidence of maximum yield and conversion but in some of the runs the maximum yield occured at a lower temperature than the maximum conversion.

Since maximum

conversion is reached at lower temperatures with longer contact times, better yields are obtained at these longer contact times.

Under these conditions high conversion and

low temperatures favor high yields.

Thus we can conclude

that the use of shorter contact times would not be desirable unless the technique of nitration can be improved to allow high conversions at lower temperatures (that is then the use of a catalyst) or to reduce by-product formation.

*3 V.

Nitration at Three Minutes Contact Time At the conclusion of the establishment of a suitable

temperature, oxygen feed, and contact time combination, it was decided to make one run which would utilize this information to set up an optimum run*

A contact time of three minutes, a

temperature of 285° 0 , and an oxygen feed of three quarters of a mole of oxygen per mole of nitrogen dioxide were used* The conversion was 2B.9‘ j> which equals the highest reproducible conversion measured in this research within the limits of error of the analytical methods used.

The yield was

which is significantly above the best previously obtained* The yields of by-products in per cent were as follows; carbon dioxide 2 1 *3 » propylene 2 *8 , ethylene 2 .8 , carbon monoxide 1.4, and aldehydes 0*3*

50. VI# By-Products Nothing has been said so far in this discussion about the variations in by-product formation with variation in conditions.

The by-product picture can be fairly

summarized for all the runs made with nitrogen dioxide#

In

general the by-products formed are predominately those of lower or no commercial value• This statement is not as critical of this method of nitration as it may seem, for the yield of nitro paraffins is so high relative to that from nitric acid nitrations as now practiced that the actual loss involved is not great. The most valuable by-products formed are aldehydes, for, besides their intrinsic value they are easily removed from recycle hydrocarbon streams. Aldehydes are not formed in large quantities in nitration under the conditions employed. In fact, in a number of runs, no measureable aldehydes was detected#

quantity of

Aldehyde formation is favored by

high temperatures, short contact times, and possibly to a slight extent by increased oxygen contact of the feed.

It

is likely that if aldehydes are formed to any extent that they decompose. This occurrence has been suggested by Addison (1), and was used by him to explain the formation of carbon monoxide in nitrations with nitric acid# The aldehyde formed predominately is formaldehyde. This is evidenced by the fact that at all but the highest

51 temperatures, all of the aldehyde is concentrated in the water layer and no measureable quantity can be extracted from the water layer by ether• By far the majority of the by-products in terms of moles of carbon contained (and even more so on a simple mole basis) consist of oxides of carbon*

The data show that the

yields of these oxides rise almost proportionately as the yields of nitro paraffins fall*

The most significant factor

in the changes in carbon oxide yields is the temperature, but in every case some oxides of carbon are obtained*

The

mole ratio of COg to 00 remained almost constant at 3/2* The only other important by-products are olefins* Propylene formation is favored by oxygen and high temperatures* Ethylene is found to a much lesser extent, but, its formation usually is favored by the same factors as those which favor the formation of propylene• These factors cause such a variation in olefin formation that it is difficult to generalize*

However, the yield of olefins is almost reduced

to the same value as the yield of aldehydes when conditions favoring high nitro paraffin yields are employed* The actual moles of aldehydes, olefins, and oxide of carbon obtained are given in the Table of the Summary for comparison*

For comparison purposes they must be adjusted to

equal amounts of hydrocarbon charged.

The differences in

hydrocarbon charged result from a number of factors including the length of the run and the size of the reactor used*

52. VII.

Yield Based Upon Nitrogen Dioxide

Throughout the early work in this study it was tacitly assumed that little or no nitrogen dioxide was reduced to nitrogen in the nitration reaction.

This

assumption was based upon the results of previous work on nitric acid nitration discussed by Addison (l).

The conditions

used in this study were much less severe than those reported. When it was discovered that as much as 20% of nitric acid was converted to nitrogen it became imperative to determine if this same conversion to nitrogen also occured in nitrogen dioxide nitrations. In run 275-2.0 an analysis of the total off gas revealed that a residual gas, presumably nitrogen, was obtained to the extent of 2 .6% of the nitrogen dioxide charged.

This method of analysis is not very accurate because

of the small volume of gas remaining. In run 275-2-1 a search for nitrogen was conducted on the so called "back gas", or gas sampled from the dry ice condenser.

The procedure is given in the section on analysis.

The presence of methane was detected and the nitrogen gas remaining corresponded to 5% of the nitrogen dioxide charged. The analysis was again handicapped by the small volume of gas remaining in spite of a six fold concentration of noncondensible gases in the dry ice condenser. It was hoped that additional information could be obtained on the question of nitrogen gas in the off gas by

538abstracting the amounts of nitrogen found as nitro paraffins, nitric oxide, and nitrogen dioxide in the off gas from the amount of nitrogen dioxide charged.

Unfortunately an accurate

nitrogen balance is not possible, because the unreacted nitrogen dioxide which originally condenses with the water and nitro paraffins can not at present be accurately measured apart from organic acids.

Thus, if this quantity is not considered,

the loss to nitrogen appears large in comparison with the determination above.

If it is considered, more nitrogen

is indicated than had been charged as nitrogen dioxide. However, the information obtained is sufficient to conclude that high nitrogen yields are to be expected in low temperature nitrations

with nitrogen dioxide.

Since nitrogen dioxide

and nitric oxide can be recycled after slight processing a loss of nitrogen of 3$ of the charge with 30% conversion amount to a nitrogen utilization of 91% - that is to say the yield based upon nitrogen dioxide is 91%*

54.

VIII.

Mechanism

The mechanism of the nitration reaction has not been completely proved, but Addison (l) has suggested a mechanism which accounts for the observed facts.

The results

of this study have not disclosed anything which violates the idea of nitrogen dioxide combining with a free radical to form either a nitro paraffin or a nitrite.

In fact the

suggestion of nitrogen dioxide being the nitrating agent accounts for the results of nitration with either nitric acid or nitrogen dioxide.

Only one question of doubt that

both methods proceed by the same mechanism has been raised. Why is nitric acid seemingly a better nitrating agent than nitrogen dioxide?

The explanation which seems most logical

is this. Eitrogen dioxide can do other things besides nitrate.

It can decompose to oxygen and nitric oxide or it

can be reduced by oxidizing the hydrocarbon.

Thus in the

first part of the reactor nitrogen dioxide is too concen­ trated and engages in too many side reactions before an adequate number of hydrocarbon free radicals can be formed. At the end of the reaction almost the opposite condition exists.

Here the nitrogen dioxide concentration is too

low to take advantage of all of the hydrocarbon free radicals present• With nitric acid the nitrogen dioxide is formed during the reaction, and those factors which cause the nitric acid to decompose can be roughly matched to the reactions which form free radicals and in this way a higher

55* percentage of the nitrogen dioxide may he utilized*

Thus

the higher efficiencies of nitric acid do not preclude the possibility of nitrogen dioxide being the nitrating agent in both cases. Another reason why nitric acid is more effective as a nitrating agent than nitrogen dioxide, may be found in the greater ease with which nitric acid can attack hydro­ carbons to produce hydrocarbon free radicals.

This it

does by decomposing into H O radicals which attack hydro­ carbons quite readily.

It would be expected on the basis

of this argument that the effectiveness of nitrogen dioxide as a nitrating agent would be greatly increased by using it in conjunction with another reagent which is capable of producing hydrocarbon free radicals at a more rapid rate.

56 IX*

Attempts to Catalyze the Nitration Reaction Many attempts have "been made to find a catalyst

for the nitration reaction.

Practically all attempts re­

sulted in finding a catalyst for oxidation rather than for nitration.

Success has been claimed but never substantiated

for oxide of antimony and arsenic (6).

It was thought that

perhaps those attempts which failed in that oxidation was increased may have resulted from the use of drastic temperature conditions and that their catalytic activity might influence nitration favorable at lower temperatures#

Therefore, for

those runs in which a comparison with non-catalytic procedure was desired, a temperature of 275°0 and a two minute contact time were used# Ferric oxide was tested as a possible catalyst by first

preparing a slurry of it in water.

This slurry

was forced through the reaction coil and then blown out with air.

This left a fairly uniform deposit of the oxide

which was dried with oxide on

a current of air#

The effects of the

a nitration are shown in Table VI#

The same general

effect was observed as had been observed in previous attempts at higher temperatures#

Poor yields and conversions were

obtained# Boric oxide catalyst was prepared by pouring a hot saturated solution of boric acid through the reactor coil#

"When the solution cooled the boric acid separated

and a fraction stuck to the side of the reaction coil#

57 • Table VI Iron Oxide Catalyst Compared to non-Oatalytic Operation

Mole Ratios

propane/nitrogen dioxide oxygen/nitrogen dioxide

Contact Time

2 minutes

Temperature

275 Catalyst

no Catalyst

Conversion

1.4$

25 . 5$

Yield

4.

62 .

4 1

By-Products (expressed as a percentage of non-catalytic operation taken at 100 $) Carbon Dioxide

312

Propylene

300

Ethylene

175

Carbon Monoxide

157

58. The remaining slurry of boric acid was forced from the coil with air. salt bath.

After air drying the coil was placed in the molten Here the boric acid decomposed to boric oxide

and steam was driven off. The results of using this reactor in a nitration run are given in Table VII.

Very little

effect was observed, but any change was in the wrong direction. When the coil was removed from the salt bath and allowed to cool it broke into many pieces and was completely destroyed. In connection with efforts to catalyze the nitration reaction the suggestion was made that if some material were added to the reaction which was more readily oxidized than propane that the reaction might be "started" or catalyzed in a homogeneous manner. For this purpose small amounts of alcohols were added to the feed in several runs.

Although

the results of these runs are not comparable in that different alcohols were tried the results are uniformly poor compared to a similar run without alcohol• This idea was tried without oxygen with the results as shown in Table VIII. It was also tried with oxygen with the results as shown in Table IX. obtained.

In no case were improved yields or conversions

59 Table VII Boric Oxide Catalyst Compared to non-Oatalytic Operation Mole Ratios Contact Time Temperature

propane/nitrogen dioxide oxygen/nitrogen dioxide

41

2 minutes 26j Catalyst

No Catalyst

Conversion

20.0

20.6

Yield

51 •

5%"

By-Products (expressed as a percentage of non-catalytic operation taken as 100 %) • Carbon Dioxide

114-

Propylene

300

Ethylene Carbon Monoxide

93 102

60 Table VIII Effects of Adding Alcohol to the Feed Without Oxygen Run - Interpolation between 246-2 and 300-1

275 -IO

Alcohol Used

none

•Pentanol

-

1

Conversion

14*9

11*5

Yield

46

45

$ of total charge

Table IX Effects of Adding; an Alcohol to the Feed With Oxygen Run Alcohol Used

275-1

275-6

275-11

none

1-Propanol

1-Pentanol

-

4

1

$ of total charge Conversion

25*5

19*7

16.4

Yield

62

61

46

6l.

EXPERIMENTAL I.

Apparatus and Technique

The apparatus used in this investigation is very similar to that used "by Addison (l) when he was employing a coil type reactor*

Thus this section will be confined to

a discussion of differences and improvements* Reactors. Three different reactors were used in this investigation.

All of these reactors consisted of two coils

of 14 mm* O.D. pyrex glass tubing*

One coil was wound on

a ten inch mandrel while the other was wound on a mandrel eight inches in diameter.

The two coils were assembled

concentrically and joined by sealing a bend between the two bottom ends*

The coils were held submerged in the salt

bath by a rack made of strap iron.

Essentially these coils

were the same except for volume. The surface to volume ratio was constant * As a matter of record the first coil had a volume of l600 ml* and was used in two initial unsuccessful runs*

The second reactor had a volume of 2000 ml. and was

used on all the remaining runs excepting 25>5-l in which the third reactor with a volume of 2200 ml* was used.

No pre­

heating was used because of the slow flow through the reactor, the low temperature, and the observed fact that the gas cooled to less than 60°0* in traveling about 3 inches of

& mm*

tubing.

This meant that when the feed tube entered

62. the reactor slightly below the liquid level of molten salt that reaction temperature had been reached since heat transfer from molten salt to glass is better than from glass to air which is the condition at the outlet. Propane and Oxygen Metering. In all experiments propane and oxygen were metered through jets in which a critical ratio of upstream to down­ stream pressure was exceeded.

This critical ratio is the

point at and beyond which the velocity of flow through the throat of the jet is that of sound at that temperature.

Above

this ratio the flow is independent of downstream pressure and is a function of upstream pressure as it affects the gas density in the throat. Originally the upstream pressure was controlled by a Hoke needle valve on the propane but after a few early experiments the pressure here as on the oxygen was controlled by a two stage pressure regulator. The jets were constructed by drawing down 8 mm. pyrex tubing to a capillary at one end. is desirable.

A long gentle taper

This jet is then ring-sealed into a piece

of larger diameter pyrex tubing.

In operation the upstream

side of the jet was connected to the pressure regulator by means of heavy rubber tubing.

An extra strong tube can be

constructed by lamination of concentric tubes just large enough to permit a smaller tube to be pulled through with a piece of heavy wire after lubrication with Silicone grease. For aliphatic hydrocarbons the inner-most piece is best if

63* made of Tygon.

Starting with 1/% x 3/S Tygon and using

two outer tubes pressure in excess of 100 pounds can "be withstood.

Secure multiple wiring of the ends to the tubes

attached if necessary. Jet size can be adjusted conveniently by making a long jet and calibrating before sealing in the final tube. If the jet is too small a small length is broken off the end and the process repeated.

The range of flow permitted

by one jet is rather small and several jets are sometimes needed.

The jets are calibrated through a wet test meter

either directly or through the apparatus.

Propane can be

and was also checked by condensing the gas with a dry ice condenser and weighing the amount collected in a given time. The rate at several pressures furnishes all the data for a calibration curve. Nitrogen Dioxide Flow. Nitrogen dioxide flow was measured by means of a differential manometer across an orfice.

The method used

for propane and oxygen could not be applied to nitrogen dioxide because sufficient pressure of nitrogen dioxide was not available.

The manometer fluid used was nitric acid

saturated with nitrogen dioxide

as used by Dor sky (4).

The

meter was calibrated by collecting liquid nitrogen dioxide in a tube sealed at one end and necked down at the other. An induction tube bringing the gas from the meter was pro­ jected down into the dry-ice cooled tube.

The meter was

6kadjusted while the nitrogen dioxide was vented to the hood through a "blow off stopcock*

Wien the time period started

the blow off stopcock was closed and the stopcock to the trap opened similtaneously• The procedure at the end of the timed period was the reverse*

The dewar was then lowered

with the trapping tube, the tube was cooled up to neck in the dry ice bath and sealed off while in this position.

After

the tube warmed up and was wiped clean it was weighed on an analytical balance.

Then a very small hole was blown in

the sealed tube by the internal pressure when a small hot flame of a hand torch was applied. out behind a shield.

This operation is carried

After the nitrogen dioxide has boiled

off the tube is reweighed.

The difference in weight is

used to calculate the nitrogen dioxide flow*

The nitrogen

dioxide meter can not be calibrated by passing the nitrogen dioxide through the reaction.

The calibration can be checked

under actual operating condition by setting the flow while it is blowing off to the hood*

Then the stopcocks may be

switched to send the nitrogen dioxide through the reactor. If the meter reading does not change the calibration is proved to be correct when flowing through the system.

In

order that this condition be fulfilled the pressure drop through the reactor must be very low.

In the case where

long contact time is used and the reactor is made of large diameter tubing the flow velocity is low and so the pressure drop is low.

This operation does not allow the use of high

65.

pressure drop mixing chambers so an ordinary tee joint is used to mix the hydrocarbon-oxygen stream with the nitrogen dioxide.

Some trouble was experienced in this tee joint in

that cold flashes were observed in some runs.

This trouble

was eliminated by placing a four liter surge vessel in the propane-oxygen line aheadof the mixing tee joint# Hitrogen dioxide is best contained in stainless steel or glass.

It attacks rubber and Tygon tubing so

that all glass tubing was used with a small piece of Tygon thickly coated with Silicone grease used tohold the glass butted up to a stainless steel nipple from the nitrogen dioxide cylinder# Product Recovery# The method of produced recovery used is identical to that described by Addison (l)# Operation. The procedure used in starting a run was as follows; (1)

The regulating valve on the propane feed

system was adjusted to the pressure required to give the desired flow# (2)

The regulating valve on the oxygen was adjusted

to the pressure required to give the desired flow. (3 ) The nitrogen dioxide regulating valve was adjusted to give the desired flow as indicated on the differential flow meter manometer.

At first the flow was

vented to the hood directly thus allowing careful adjustment before the run time was begun.

66.

(11.) when the nitrogen dioxide flow was steady the venting stopcock was closed and at the same time the stopcock to the nitration system was opened.

At this time

the run timer was begun* (5 )

After a duration of time equal to the contact

time had passed the gas coming from the product recovery flask immediately dowstream from the water condenser was switched from venting to the hood to passing through the dry ice condenser » During the run it was necessary to adjust the flow of nitrogen dioxide several times.

In the course of

the long runs, necessitated by long contact times and limited reactor volume, the nitrogen dioxide regulating valve showed a tendency to creep shut• Gas samples were taken during the run. After sufficient material had been collected the run was terminated in the following manner: (1)

The stopcock between the nitrogen dioxide

flowmeter and the

nitration system was closed and at the

same time the venting stopcock was opened. This concluded the run time. (2)

The nitrogen dioxide regulating valve

(3)

The propane flow was increased to hold the

was shut•

contact time for the mixture already in the reactor constant.

67. (4)

At the end of the duration of run time plus

a time equal to the contact time the gas stream was diverted to the hood. (5 ) The oxygen flow was stopped* (6 ) The propane flow was stopped* After shuting down in this manner the products collected by the water condenser and the dry ice condenser were weighed, and the wet test meter reading and temperature were recorded*

68. II.

Product Analysis

The following methods of product analysis were worked out in conjunction with Mr. L. M. Addison. Nitroparaffins. In order to use a method of analysis for nitroparaffins substantially free from possible methodic errors the analyses for nitroparaffins in this study were made by one of two modifications of a Kjeldahl procedure proposed by Somers

(9)«

The analysis by fractional distillation

as performed by previous workers suffers from such diffi­ culties as inclusion of water and oxygenated by-products in the nitroparaffin cuts, and the necessity of basing the molecular weight of the nitroparaffin on fractionation data. In Dor sky *s (4-) work the samples produced in each run were so small that the products from a large number of runs had to be combined and one fractionation made#

The results of

each run were based upon the fraction contributed to the distillation charge.

This assumes that the various changes

in variables had no change in product composition. To prepare the crude product for Kjeldahl analysis the following scheme was used.

The dry ice condensed liquid

was allowed to slowly boil away as it warmed up to room temperature.

The residue from this operation was combined

with the product condensed by the water cooled condenser. This total product was extracted repeatedly with ether to remove all the nitro paraffins present.

The ether layer

69 •

was treated with solid sodium 'bicarbonate to remove acidic material since a part of this material might have been nitrogen acids• A stronger base would cause condensations between the nitro paraffins and any aldehydes present.

An

aliquot portion of the neutralized extract was diluted with ethyl alcohol to the mark on a small volumetric flask*

70. Improved Ejeldaîil Analysis for Nitro Compounds A sample of the alcoholic solution prepared to have about 3 mg. of nitrogen in each two ml. of solution is pipetted into a 100 ml. Kjeldahl flask containing 1 ml. of reagent (20 per cent solution in concentrated hydro­ chloric acid) • Two glass beads are added and the solution is boiled for 2 minutes.

After cooling, 0.5 g. KgSO^, 50 mg.

OuSO^, 50 mg. HgO, and 5 ml. of concentrated H^SOij. are added and the mixture is digested until it is white or a very light gray if digestion has proceeded four hours.

Extreme

care must be taken to avoid foaming at the beginning of the digestion.

After digestion the mixture is diluted with

20 ml. of water in such a way as to wash the neck of the flask. The flask is placed on the micro Kjeldahl apparatus described by Sommers (9)*

The aspirator is turned on, the flask

heated for 2 minutes.

The delivery tubes and condenser are

then washed down before the receiver containing the solution to be titrated is fitted to the apparatus.

This solution

is about 30 ml. of 2?fo boric acid and serves to absorb the ammonia distilled off.

The Kj eldahl flask is cooled in an

ice bath before 50$ aqueous NaOH containing 2.5 g. of NagS per liter is added in slight excess of the amount necessary to turn the contents of the flask a dirty brown color.

The

ice bath is removed and heat is applied by a micro burner placed several inches below the flask.

Heating is continued

for 15 minutes • After this the condenser and delivery tube are washed down into the receiver, the contents of which are titrated directly with 0.01 H. sulfuric acid.

A mixed

creed green-methyl red indicator described by Somers

(9)

is used and the end point is slightly beyond a pale green almost colorless solution to the faintest pink tinge detectable over a. white paper* ,v

72 Amine Method for Lower Nitro Paraffins The digestion step in the above method is trouble some and slow*

It is omitted in the so called amine method

since the amines formed hy reduction of the lower nitro paraffins are volatile and even more basic than ammonia» This method gives more precise results with greater speed and simplicity than the digestion method. The procedure is the same up to the digestion step*

After the reduction with T^Ol^ the flask is filled

with 20 ml* of water and placed on the apparatus*

The

receiver containing the 2^ boric acid is put in place and about 15 ml* of 30% sodium hydroxide solution is added to the distilling flask*

The mixture is distilled 15 minutes

as before and titrated as before*

At the end of the

distillation the black color formed upon addition of the base should be completely gone and the contents a pure white*

It is sometimes necessary to add a little distilled

water during the distillation or to shake the apparatus to remove solid dark colored material from the sides of the flask just above the liquid level*

73.

Aoldio Materials» The process described for product analysis results in an aqueous raffinate and an ether extract.

An aliquot

portion of each (before neutralization in the case of the extract) is titrated with standard base to both the methyl orange and phenolphthalein endpoints.

These data permit

the calculation of total moles of weak acid and strong acid. The usefulness of these results is somewhat open to questions. Some weak inorganic acid (e.g. nitrous acid) was undoubtedly present. However the method demonstrated that the amount of carboxylic acid formed was not great•

74*

Carbonyl OompouncLs» An aliquot portion of water and ether layer was analyzed for carbonyl compounds in conjunction with the determination of acidity to methyl orange• After a methyl orange endpoint was reached a few milliliters of a saturated solution of hydroxylamine hydrochloride was added. After a few minutes shaking the hydrochloric acid liberated by reaction of hydroxylamine hydrochloride with aldehydes and ketones was titrated back to the original methyl orange endpoint. The hydroxylamine hydrochloride solution itself is adjusted to a pH of 3-4 by titrating an aliquot portion in a blank titration with 0*5 H standard base.

The cal­

culated amount of base for the total solution of hydroxyl­ amine hydrochloride is added to the stock bottle.

75 Pas Analysis^ An Orsat type gas analysis was built which was capable of analysis for nitrogen dioxide, carbon dioxide, nitric oxide, propylene, ethylene, oxygen, carbon monoxide, propane, methane, hydrogen, and nitrogen.

Because of the

time comsuming nature of the process the last four materials were determined in only a few cases*

Gas samples were

introduced into this apparatus from gallon jugs where the gas was collected by displacement of a saturated solution of sodium sulfate made acid to methyl orange with sulfuric acid.

The gases were removed in the order listed below

where the reagent used is described* (l)

Nitrogen dioxide.

Absorbed in 50$ sulfuric

acid using a fresh 1 ml. portion of the absorbent for each pass.

This was carried out in a special 125 ml. pipette

packed with glass tubing.

The top of the pipette is

sealed to a three way oblique bore stopcock.

This stopcock

allows the pipette to be connected to either the manifold to which it is sealed or a second three way stopcock* When the first stopcock is turned to connect the pipette to the second stopcock it is in position to remove or reject the sulfuric acid.

This choice is determined by

the setting of the second stopcock which is connected to an acid discharge tube and an acid reservoir.

The acid

reservoir consists of a calibrated centrifuge tube.

Acid

and gas were admitted or discharged by raising or lowering

76 a mercury leveling bulb connected to the bottom of the pipette*

The leveling bulb supplies mercury to fill the

pipette prior to addition of acid and the sample * (2) KOH solution*

Carbon dioxide*

Reacted with y^jo aqueous

If the base was not exhausted it will remove

all the carbon dioxide in three passes*

This fact was

checked by repeating the operation to see if a constant reading was obtained* (3 ) Nitric Oxide• Absorbed in a saturated solutbn of ferrous sulfate in 15% sulfuric acid until a constant reading was obtained. (4)

Propylene absorbed in 61$ and 66$ sulfuric

acid contained in the pipette described previously under nitrogen dioxide.

Fresh 1 ml* portions of 61$ acid were

used until absorption per pass was less than 0*5 ml. Eighty-eight per cent acid was then used until the loss per pass was constant.

The constant loss times the number of

passes made with 66$ acid subtracted from the total loss in both strengths of acid gave the quantity of propylene present* (5 ) Ethylene*

Absorbed in 96$ sulfuric acid

containing 1$ silver sulfate in the same manner as with the weaker acid*

The constant absorption times the number

of passes is subtracted from the total loss in 96 $ acid* The constant absorption in 66$ acid times the number of passes made with this acid are added to this figure to arrive at the quantity of ethylene.

77 This may be restated in the form of an equation: Volume of ethylene = Loss in 96% acid - constant loss with 96 $ acid x the number of passes made with 9 ^ acid -f constant

loss in

acid x the number of passes made with this acid.

This corrects for ethylene previously absorbed with propylene and also for the propane absorbed with the ethylene. (6)

Oxygen absorbed in caustic pyrogallol to a

constant reading.

This solution was prepared as follows:

Fifty grams of pure pyrogallic acid was dissolved in 1^0 ml. of distilled water.

Twelve hundred grams of pure potassium

hydroxide was dissolved in 800 ml. of distilled water. potassium hydroxide solution was cooled.

The

The two solutions

were mixed in a large bottle protected from the air by a tight rubber stopper. for oxygen frequently.

It is advisable to change the solution Even if the solution still absorbs

oxygen, an old solution may lead to errors caused by the evolution of carbon monoxide. (7 ) Carbon monoxide.

Ammonical cuprous chloride

solution is used as the absorbent to a constant reading. This solution is prepared by dissolving 200 g. of cuprous chloride in 750 ml. of distilled water.

This is allowed to

stand for 24- hours in a tightly stoppered bottle containing 50 g. of Ou turnings or mesh.

For use in the pipette 100 ml.

of the stock solution is mixed with 36 ml. of concentrated aqueous ammonia of specific gravity O.89 . The pipette also contains copper turnings to the extent of 10 g.

73. (5)

Propane• Kerosene is used as the absorbent *

The bulk of the propane is absorbed in the first pipette full of kerosene but complete removal requires that the kerosene be renewed once# (9)

Hydrogen and methane are calculated from the

loss of volume following the burning of the sample mixed with excess oxygen and the loss of volume when the carbon dioxide formed in the combustion process is removed in the pipette previously described under carbon dioxide# burning is carried out over hot copper oxide#

The

The volume

of methane is equal to the loss of volume in the carbon dioxide pipette#

The volume of hydrogen is equal to two

thirds the contraction on combustion minus four thirds the carbon dioxide volume found# (10)

Nitrogen#

Obtained by difference of gas

sample before adding oxygen minus the sum of the volume of hydrogen and methane. Minor errors are encountered if the potassium hydroxide, ferrous sulfate, alkaline pyrogallol, and cuprous chloride solutions are not saturated with propane before the sample is analyzed.

Errors introduced in this manner are

more important if a more soluble hydrocarbon such as butane is being nitrated#

III.

Calculations

Contact Time. The contact time was calculated on the basis of reactor inlet conditions.

The weights or volumes of the

materials charged to the reactor are converted to moles per minute.

Then the volume this amount would occupy under the

reaction conditions is calculated by means of the gas laws. This volume is divided into the volume of the reactor giving the contact time in minutes.

If the contact time in seconds

is desired, the same process is repeated using the total moles per second charged to the reactor. The calculation of the contact time in this manner is only correct if there is no change in the number of moles during the reaction.

This condition is approached because

the net change in number of moles is not great• Moreover the change from one run to another is very small so that the data are comparable and mutually consistent. Conversion. The conversion referred to is that based on the percentage of nitrogen dioxide charged which appears in the form of nitro paraffins. Moles of Products. Actual quantities determined have been reported in most cases, but in some cases the moles reported have been corrected for imperfections in the material balance. Ho such corrections were made if the material balance was drastically off or in any case where the conversion was at

go or near a peak.

For comparison purposes between runs the

molar quantities must be adjusted to equal amounts of hydro­ carbon charged as has been discussed under By-Products. Yields. The yield considered is the yield of nitro paraffins based upon propane.

The number of carbon atoms in the mole­

cules of the by-products has been taken into account. Thus one mole of propane can yield three moles of carbon dioxide but only one and one half moles of ethylene.

All yields are

based upon a percentage of the amount of raw material which is changed from its original form and can not be recycled.

IV.

Chemicals Used

Propane. Phillips Petroleum Company1s Technical Grade propane was used.

This material is guaranteed to be better

than 95 mole per cent pure* Oxygen. Pure oxygen was used as supplied by Linde Air Products Company. Uitrogen Dioxide. Matheson Chemical Company1s pure grade was used direct from the cylinder. Nitric Acid. Baker's O.P. Analyzed Grade was used.

The assay

was confirmed on every bottle used by titration with standard base. 1-Propanol. DuPont1s commercial grade 1-Propanol was used. 1-Pentanol. Columbia Chemical Company's 1-Pentanol was used.

82 SUMMARY (1)

It has "been demonstrated that high yields

"based upon hydrocarbon charged are possible in nitrations with nitrogen dioxide• Values up to 71% have been achieved* (2)

It has been demonstrated that nitrations

with nitrogen dioxide can be carried out with high yields based upon nitrogen dioxide*

In one case only 3% of the

nitrogen dioxide was lost from possible recycle in the form of nitrogen gas* (3)

It has been demonstrated that the best

previous conversion based upon nitrogen dioxide of 26% can be raised to 29% while the contact time is reduced from l4 minutes to 2 minutes by the use of oxygen and selected operating temperatures* (4)

It has been demonstrated that high conversions

and yields on both nitrogen dioxide and propane are achieved simultaneously* (5)

The effect of oxygen on by-product formation

is a general increase in the amount of by-products produced rather than a pronounced shift in the relative distribution of these by-products*

Oxygen increases the rate of formation

of nitro paraffins faster than it does the by-products*

This

accounts for the increase in yield with the use of oxygen* (6 ) Formaldehyde is substantially the only carbonyl compound formed in the nitration reaction using nitrogen dioxide as the nitrating agent*

f53*

(7 ) Increases in temperature favor the formation of hydrocarbon free radicals necessary for the nitration pro­ cess hut tend to destroy nitrogen dioxide.

Therefore optimum

temperatures are observed above and below which the yields are lower. (8 ) If a method of forming free radicals at low temperature could be found and used higher yields and conversions would result. (9) Hitrogn dioxide is the nitrating agent. such

it is all charged in the reactor inlet.

As

The formation

of hydrocarbon free radicals proceeds throughout the reactor. This mismatch of the two radicals whose combination produces nitro paraffins is a serious fault in the present technique of nitrogen dioxide nitrations. (10) Addisonfs mechanism is substantiated for nitrations with nitrogen dioxide.

BIBLIOGRAPHY Addison, L. M., A Study of the Vapor Phase Nitration of Butane with Emphasis on the Effect of Adding Oxygen, Ph. D. Thesis, Purdue University (1950) • Alexander, Loyld, A Study of the Vapor Phase Nitration of Paraffins, Ph. D. Thesis, Purdue University (19^7)• DeVries, Thos., and Rodehush, W. H., The Thermal Dissociation of Iodine and Bromine, J. Am. Ohem. Soe., 656-65 (1927 ). Dorsky, Julius, Vapor Phase Nitration of Propane with Nitrogen Dioxide, Ph. D. Thesis, Purdue University (1940)

.

Guyer, A., and Rufer, A., Bromination of Propane, Helv• Chim. Acta., 2^, 533-41 (19*10). Kistiakowsky, G. B., and Van Artsdalen, E. R., Bromination of Hydrocarbons I. Photochemical and Thermal Bromination of Methane and Methyl Bromine, J. Chem. Phys., 12, 469-?S (1944). Levy, Norman, Brit. Pat. 527 ,0 3 1 , Oct. 1, 1940. Levy, Norman, (To Imperial Chemical Co.) 2,3^2,241, Aug. 14, 1945. Levy, Norman, (To Imperial Chemical Co.) 2,394,315» Feb. 5 , 1946. Perelis, W. J., Bromination of Saturated Aliphatic Hydrocarbon Gases, Ind. Eng. Chem., 25., 1160-1 (1933) • Riley, Elizabeth F., Private Communication, (1949)• Somers, P. D. Jr., Proc• Indiana Acad. Soi., 54, 117-20 (1944).

8

$ .

SUMMARY OF PRIMARY DATA FOR NITRATIONS WITH NITROGEN DIOXIDE (Runs with Proposed Catalyses Omitted) 24S-la

300 -ib

300-2

248-2

300-3

325-1

Temp., °C.

246

300

300

248

300

325

Run Time,Min.

267

360

270

220

180

235

Propane

4.560

—

4.720

4.120

2.790

3-950

Nitrogen Dioxide

1.100

——

1.130

1.100

O.63O

0.940

Oxygen

0 .0

——

0 .0

0 .0

0.321

0 .0

0.092

-------

0.162

0.151

0.142

0 .1 5 6

—

-------

0.190

0.353

0.071

0 .0 6 6

—

-------

0.095

0.033

0.031

0.0 91

Run

Moles Charged

Moles Products RNO2 cvi

0 0

°3H6 CA CO

—

-------

0.000

0.000

0.007

0,000

—

——

0.095

0.534

0.132

0.1 1 2

Strong Acid

—

——

——

-------

"

—

Total Acid

— -

-------

—

-------

NO

—

-------

0.390

0.340

—

0.159

0.534

A

2 Aldehydes

"

0.007

#6,

Run

325-2

325-3

300-4-°

300-5

275-1

275-2

Temp., °0,

325

325

300

300

275

275

Run Time,Min.

120

2^5

120

120

215

240

Propane

1.301

3*200

3.500

1.300

1.520

3 .3 &O

Nitrogen Dioxide

O.3O5

0.735

0.320

0.4-31

0.336

0.770

Oxygen

0 .1 5 4

O .705

0.035

0.4-22

0 .4-4-9

0.4-30

0 .0 6 1

0.1 00

0.135

0.071

0.099

0.14-5

ro

Moles Charged

0 .0 7 4

0.122

—

0.091

0.031

0.1 63

°3h6

0.0 27

0.0 1 6

—

0.003

0.002

0 .0 3 6

0.002

0 .0 3 6

— —

0.019

0.019

0 .0

00

0.031

0.204-

—

0.069

0.055

0.0 6 1

Strong Acid

-

—

—

—

—-

—

Total Acid

-

—~

—

—

——

NO

0.214

0.493

—

0.066

0.00

-

°2 Aldehydes

——

——

—

—

-

0.00

——

0.002

Trace

Moles Products

8

rno2

—

—

ST-

255-1

300-6

275-3

245-1

255-2

255-3

Temp., °0.

255

300

275

245

255

255

Run Time,Min.

180

360

355

330

300

250

Propane

3.040

2.710

2.660

2.540

2.730

2.0 60

Mtrogen Dioxide

O .711

0 .7^8

0.723

0.687

0.607

0 .4 7 9

Oxygen

0.384

0.3 5 4

0 .34 g

0.3 16

0.309

0.453

r no 2

0.092

0.122

0.131

0.1 16

0.131

0.108

C02

0.094

0.140

0.112

0.072

0.093

0.226

0.009

0.031

0.010

0.009

0.018

0.002

Run

Holes Changed

Moles Products

°3H6 c2 % 00

0.004

0.003

0.012

0.000

0.000

0.000

O.O56

0.1 06

0.100

0 .0 7 8

0.1 1 7

0.1 05

Strong Acid

—

—

—

—

—

—

Total Acid

-

——

—

—

—

HO

0.00

0.284

0.177

0.009

0.000

0.000

—

—

-

—

-

—

— —

0.006

Trace

—

-

—

°2 Aldehydes

—

—

—

33.

275-4°

255-4

300 -7 °

275-5

325-4

Temp., °0.

2?5

255

300

275

325

275

Run Time,Min.

253

120

120

120

120

146

Propane

1.360

2.020

2.510

2.060

2.120

1.590

Nitrogen Dioxide

0.413

0.460

0 .5 5 3

0.577

O.55 S

0.5 3 0

Oxygen

0.455

0.474

0 .6 1 4

0.623

0.590

O .536

rn o 2

0.135

0.062

0.144

0.1 06

0.0 9 6

0.104

002

0 .32 s

0 .10 S

0 .IS3

0.162

0.147

0.086

%

0.002

0.002

0.020

0.031

0.0 3 2

0 .0 0 9

0.000

0.000

0.020

0.000

0.012

0.000

CO

0.2 27

0.196

O .165

0.036

0.1 2 3

0.086

Strong Acid

0.033

—

—

0.005

0.010

— —

Total Acid

0.035

0.012

—

NO

0.000

Run

275-6

Moles Charged

Moles Products

°2 Aldehydes

0.000

0.063

0.000

0.132

0.007 0.000

0.0 07

-

0.000

0.003

0.000

0.003

0.000

—

0.001

0.044

59»

tun

300-8

275-7

275-8

275-9

275-10

275-11

300-9

Tempe, °0.

300

275

—

275

275

275

300

Run Time,Min.

252

270

—

239

180

185

120

Propane

1.870

3.360

—

2.14-30 2.850

2.680

2.250

Nitrogen Dioxide

O.lj-51 0.4-61

— —

OA77

0 .6 9 6

O .655

O .563

Oxygen

0.4-70

0.4-84-

—

0.5314- 0.7 30

0.686

O.55 O

rno2

0.102

O .127

—

0.133

0.080

0.120

0.130

C02

O.I34- O .151

—

0.130

0.108

°3h6

0.014- 0.00-4 —

0.003

0.025

0.015

0.014

0.014- o.oo4

——

0.003

0.007

0.015

0.014

CO

0.084- 0.025

——

0.072

0.082

0.118

0.0 59

Strong Acid

0.018

0.012

— —

— —

0.004

0.0 0 6

0.0 07

Total Acid

0.019

0.0 17

— -

0.010

0.010

0.008

NO

0.061

0.0 00

—

0.030

O.57 O

0.000

0.000

°2 Aldehydes

0.000

0 .0 32

— —

0.000

0.000

0.170

0.081

0.013

0.0 02

—

0.009

0.0 13

0.001

0.001

loles Charged

loles Products

g 2h4

0.173

a.

Gas sample lost•

b*

Isolation procedure resulted in condensation of nitro paraffins and aldehydes•

c* Poor material balance. d . Irregular Nitrogen Dioxide Flow.

0.1 6 8

VITA

James Veith Hewett was born July 25 , 1921» in Broken Bow, Nebraska, graduated from Hastings High School, Hastings, Nebraska, in 1939> received the B.S. degree from the University of Nebraska in 194-3 • He was employed by Phillips Petroleum Company from 194-3 until 194-6 when he was given a leave of absence to attend Purdue University. He received the M.S. degree in 194-8.