CASE FILE - NASA Report CASE FILE lrPYIh COW! .d November 1970 ... those for the fluorine-caustic...

AN ECONOMIC STUDY OF' OXYGEN DIFXUORIDE

Final Report C A S E FILE lrPYIh C O W ! .d

November 1970

National Aeronautics and Space Administration Headquarters Contracts Division

Washington D. C.

(Prepared under Contract No. NASW-1911

Allentown, Pennsylvania by Air Products and Chemicals, Inc.,

F. L. man and J. F. Tompkins, authors)

https://ntrs.nasa.gov/search.jsp?R=19710010791 2018-06-15T14:56:06+00:00Z

I

AN ECONOMIC STUDY OF OXYGEN DIFLUORIDE

F. L. Hyman J. F . Tompkins

A i r Products and Chemicals, Inc. Allentown, Pennsylvania

FOREWORD

This i s t h e f i n a l report on NASA Contract NASW 1911 "An Economic Study of Oxygen Difluoride". from June 4, 1969 t o June 4, 1970. i s t o inves t iga te t h e economics of processes f o r t h e preparation of oxygen d i f luo r i ae , and t o specify a process o r processes by which it may be supplied below the present market pr ice . oxygen di f luor ide w i l l be needed a t uneven rates with 2,000 t o 4,000 lbs required per year. manager t h i s f i n a l report presents economics f o r producing OF2 at 4,000 t o 50,000 lb /yr . t h e possible use of OF2 fo r engine t e s t ing as well as ac tua l mission requirements.

The work reported covers t h e period The object ive of t h e contract

The

A t t h e request of t h e NASA technical

This increased production i s based on

The economics presented are based on literature y ie lds except those for t h e fluorine-caustic process which a re based on experimental work performed under t h i s contract . Some l imited experimental work w a s a l s o performed on the porous anode e l e c t r o l y t i c method.

The NASA Program Manager f o r t h i s contract i s D r . Robert S. Levine, NASA Headquarters, Washington, D. C. and the Technical Manager is Mr. Wolfgang Simon, J e t Propulsion Laboratory, Pasadena, California.

i

ABSTRACT

The economics of owgen dif luoride production based on four processes described i n t h e l i t e r a t u r e are presented. These four processes are: 1) fluorine-caustic 2 ) fluorine-water catalyzed by a l k a l i f luorides and anode. processes are reviewed. It is concluded t h a t at t h i s t i m e t h e fluorine-caustic o f f e r s the least expensive method f o r making OF2 at rates at or below 10,000 lb/yr . are the r e l a t ive ly l aw process development costs and reasonably w e l l defined y ie lds and investment costs.

3) e lec t ro lys i s of water i n hydrogen f luoride, 4) e lec t ro lys i s of oxygen i n hydrogen f luoride using a porous

The literature f o r these four processs and other po ten t i a l

The main reasons f o r t h i s choice

ii

TABLE OF CONTENTS

Page I . S.ary . . . . . . . . . . . . . . . . . . . . . . . . . . 1 I1 . Introduction . . . . . . . . . . . . . . . . . . . . . . . 4

111 . Literature Review . . . . . . . . . . . . . . . . . . . . . 6

A . Reactions of Fluorine with Inorganics . . . . . . . . . 6

1 . Fluorine-Caustic . . . . . . . . . . . . . . . . . 6

2 . Fluorine-Water Reactions . . . . . . . . . . . . . 7 3 . Other Fluorine-Inorganic Reactions . . . . . . . . 10

B . Direct Combination of Oxygen and Fluorine . . . . . . . 10 C . Direct Electrolysis . . . . . . . . . . . . . . . . . . 11

1 . WaterinHF . . . . . . . . . . . . . . . . . . . . 11 2 . Water and Organic Compounds . . . . . . . . . . . . 13 3 . Direct Electrolysis of Inorganic Compounds . . 14

4 . Direct Electrolysis Using Oxygen . . . . . . . . . 14 IV . OF2 Production Economics . . . . . . . . . . . . . . . . . 17

A . Basic Assumptions . . . . . . . . . . . . . . . . . . . 17 B . Process Comparisons . . . . . . . . . . . . . . . . . . 18

1 . Fluorine-Caustic . . . . . . . . . . . . . . . . . 18 2 . Fluorine-Water (Alkali Fluoride Catalyzed) . . . . 19 3 . Electrolysis of Water in HF . . . . . . . . . . . . 19 4 . Electrolysis Using Oxygen in HF with a Porous

Anode . . . . . . . . . . . . . . . . . . . . . . . 19 C . Factors Affecting the Selling Price . . . . . . . . . . 20

1 . Design Criteria . . . . . . . . . . . . . . . . . . 20

2 . Investment Factors . . . . . . . . . . . . . . . . 20

iii

.

TABLE OF CONTENTS (Continued)

Page

3 . Raw Materials . . . . . . . . . . . . . . . . . . . 26

4 . Purif icat ion . . . . . . . . . . . . . . . . . . . 26

D . Delivered Price . . . . . . . . . . . . . . . . . . . . 26

1 . Cylinder Investment . . . . . . . . . . . . . . . . 28

2 . Shipping and Storage Inventory . . . . . . . . . . 28

3 . Transportation Costs . . . . . . . . . . . . . . . 28

E . Production Lead Time . . . . . . . . . . . . . . . . . 32

F . Inf la t ion Factors .................... 32

G . Multiple Y e a r Contracts . . . . . . . . . . . . . . . . 32

V . Experimental . . . . . . . . . . . . . . . . . . . . . . . 35

A . Fluorine-Caustic System . . . . . . . . . . . . . . . . 35

Experiments at 0.1 and 0.2 lb F2/hr . . . . . . . . 35

a . Experimental Procedures . . . . . . . . . . . . 35

1 .

b . Experimental Di f f icu l t ies . . . . . . . . . . . 38

c . Operating Variables . . . . . . . . . . . . . . 40

d . Discussion of Results . . . . . . . . . . . . . 42

2 . Experiments a t 2.0 lb F2/hr . . . . . . . . . . . . 47

a . Procedures . . . . . . . . . . . . . . . . . . 47

b . Results .................... 47

c . Pur i f ica t ion . . . . . . . . . . . . . . . . . . 47

B . The Elec t ro ly t ic Oxygen Porous Anode System . . . . 49

1 . Procedure . . . . . . . . . . . . . . . . . . . . . 49

2 . Results . . . . . . . . . . . . . . . . . . . . . . 49

C . Analytical Procedure and Equipment . . . . . . . . . . 51

i v

TABLE OF CONTENTS (Continued)

Page

VI . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . 52

V I 1 . Bibliography . . . . . . . . . . . . . . . . . . . . . . . . 53

A . Fluorine-Caustic References . . . . . . . . . . . . . . 53 B . Fluorine Oxygen References . . . . . . . . . . . . . . . 54

C . Direct Electrolysis References . . . . . . . . . . . . . 55 D . General References . . . . . . . . . . . . . . . . . . . 56

t

LIST OF TABLES

Page

I - Comparison of Li terature Data for OF2 by Fluorine- Caustic Reacxions. . . . . * . . . . . . . . . . . . . . . 8

I1 - Comparison of Li terature Data for OF2 by Direct Electrolysis . . . . . . . . . . . . . . . . . . . . . . . 16

I11 - Design Criteria fo r Various OF2 Processes. . . . . . . . . 21 I V - Investments f o r Various OF2 Production Processes-

10,000 l b OF2/12 month Contract. . . . . . . . . . . . . . 22

V - Investments f o r Various OF2 Production Processes- 20,000 l b 0F2/12 month Contract. . . . . . . . . . . . . . 23

V I - Comparison of Required Raw Materials f o r Various OF2 Processes. . . . . . . . . . . . . . . . . . . 27

V I 1 - Estimated Delivered Price fo r OF2 - Single 12 Month Contract . . . . . . . . . . . . . . . . . . . . . . 29

V I 1 1 - Estimated Delivered Price f o r OF2 on a Second C o n t r a c t . . . . . . . . . . . . . . . . . . . . . . . . . 31

I X - Experimental OF2 Yield Data Using the Fluorine- Caustic System . . . . . . . . . * . . . . . . . . . . . . 41

X - Summary of Experimental OF2 Results by Caustic - Fluorine Method October, 1969. . . . . . . . . . . . . . . 43

X I - Summary of Experimental OF2 Results by Caustic - Fluorine Method November, 1969 . . . . . . . . . . . . . . 44

X I 1 - Summary of Experimental OF2 Results by Caustic- Fluorine Method December 1969. . . . . . . . . . . . . . . 45

X I 1 1 - Experimental OF2 Data Using the 2 lb /h r Fluorine- Caustic System . . . . . . . . . . . . . . . . . . . . . . 48

vi

LIST OF FIGURES

Page

I - Selling Price for OF2 at Various Production Levels for Different Process. . . . . . . . . . . . . . . . . . . 2

I1 - Selling Price for OF2 at Various Production Levels by Fluorine-Caustic and Electrolysis of Oxygen in HF Processes for a Depreciated Plant with no Process Development Expense. . . . . . . . . . . . . . . . . . . . 3

I11 - Schematic Breakdown of the OF2 Processes . . . . . . . . . 24 IV - Cumulative Selling Price for OF2 at Various Total

Production Quanities by the F2-Caustic Process fo r Various Contract Periods . . . . . . . . . . . . . . . . . 33

V - Cumulative Selling Price of OF at Various Production Levels by the Electrolysis of 8xygen in HF for Various Contract Periods . . . . . . . . . . . . . . . . . . . . . 34

VI - Fluorine-Caustic Experimental System for Producing OF2 . . 36

VI1 - OF2 Explosion Chamber. . . . . . . . . . . . . . . . . . . 37

VI11 - OF2 0.1lb F2/hr Reactor . . . . . . . . . . . . . . . . . 37

IX - Electrolytic Cell Experimental System for OF2. . . 50

vi i

I. summary

This report presents t he economics of owgen dif luoride production based on four processes described i n t h e literature. These four processes are: 1) fluorine-caustic 2) fluorine- water catalyzed by a lka l i f luorides 3 ) e lec t ro lys i s of water i n hydrogen f luoride, and f luoride using a porous anode. The literature f o r these four processes and other po ten t ia l processes are reviewed.

4 ) e lec t ro lys i s of oxygen i n hydrogen

The economic advantages and disadvantages of t he processes are discussed together with the important cr i ter ia t o be considered i n se lec t ing a process. shipping and storage costs are discussed f o r all processes at production rates of 4,000, 10,000 and 20,000 lb /yr , and at 50,000 lb /yr f o r t he fluorine-caustic and e l e c t r o l y t i c porous anode processes.

Investment, process development,

Experimental results are presented for t h e fluorine-caustic and the e l ec t ro ly t i c porous anode process. obtained f o r the fluorine-caustic process a t f luorine rates of 2 lb/hr .

Yields of 70% o r b e t t e r were

It i s conclud.ed t h a t at t h i s t i m e t h e f luorine-caustic of fe rs t h e least expensive method f o r making OF2 at rates at or below 10,000 lb/yr . The main reasons f o r t h i s choice are the rela- t i v e l y low process development costs and reasonably defined investment costs. The more economical process f o r high production rates (50,000 l b / y r ) , appears t o be t h e e l e c t r o l y t i c oqgen porous anode process i f t h e l i t e r a t u r e yields can be maintained on scale-up and estimated development costs prove r e a l i s t i c .

Estimated s e l l i n g pr ices f o r OF2 produced f o r all four processes are given i n Figure I up t o 20,000 lb/yr . l i n g pr ice fo r t h e fluorine-caustic and e l e c t r o l y t i c porous anode processes are given i n the range of 4,000 t o 50,000 lb /y r i n Figure I. These pr ices are for a single 12 month continuous production contract and include writ ing off a l l process development and OF2 production and purif icat ion investment costs over t h e 12 month contract . Estimated se l l ing pr ices f o r OF2 after process development and OF2 production and pur i f ica t ion equipment have been fully depreciated are shown i n Figure 11. For t h i s depreciated investment case the lower operating costs of the porous anode e l ec t ro ly t i c process become s igni f icant and result i n a lower s e l l i n g price. The estimated OF2 s e l l i n g pr ices do not include storage, transportation and working cap i t a l costs.

I n addition t h e sel-

1

Plant Depreciation and Development Cost Charged Over a Single 1 2 Month Contract

No Storage, Transportation o r Working Capit Charges Included

OF2 Stored i n Government Owned Cylinders up t o 20,000 Lbs.

OF2 Stored i n Government Owned Liquid Tra i l e r Above 20,000 Lbs.

l i i i i i l i i l i i i i i i i i i i i i i i i i i i i i i i i i i i l ~ i i i i i i i i i i i l i i i i i i i i i i i i l i i i l r i i l 2

1 1 1 1 1 1 1 1 1 1 1 1 1

1 1 1 1 1 1 1 1 1 1 1 1 1 I I I I I I I I I I I I 1

FIGURE I1

Sel l ing Price for OF2 at Various Production Levels by Fluorine-Caustic and

No Storage, Transportation and Working Charges Included.

Electrolysis of Oxygen i n €IF Processes for a Depreciated

Plant with no Process Development Expense

Capital

OF2 Stored i n Government Owned Cylinders up t o 20,000 lbs

OF2 Stored i n Government Owned Liquid Trailers Above 20,000 lbs

Prices Assume Continuous Production

3

11. Introduction

The object of t h i s contract w a s t o f i n d t h e most economical process f o r producing oxygen d i f luor ide at uneven rates with 2,000 t o 4,000 l b s required per year. A t the request of t h e NASA Technical Manager the program objectives were broadened i n September 1969 t o include; accelerated experimental work on the most ea s i ly reduced t o prac t ice type process, and t h e invest igat ion of t h e economics of a l l promising processes at production rates up t o 50,000 lbs/yr . requested t h a t t h e three major production rates considered be 10,000 lb /y r , 4,000 lb /yr , and 20,000 lb/yr with fluorine-caustic and porous anode e l ec t ro ly t i c processes invest igated up t o 50,000 lb/yr . portance.

For t h i s f i n a l report t h e Technical Manager

The rates are given i n order of decreasing im-

The four processes considered i n t h i s report as having poten t ia l fo r producing OF2 i n commercial quant i t ies are:

1. Fluorine-Caustic: Defined by Ruff and Meiizel (A-14) i n 1930 later ref ined by others and most recent ly studied by Brown (A-2) i n 1968. The process consis ts of bubbling f luor ine through 2% caust ic sclut ion. The f luorine i s converted t o OF2 and an a l k a l i f luoride as defined by the following equations:

2 NaOH + 2 F 2 r 2 BaF + OF2 + H20

2. Fluorine + Water: Borning and Pullen (A-3) modified t h e fluorine-water reaction by running it i n the presence of an a l k a l i f luoride.

3. Direct e l ec t ro lys i s of Water i n Hydrogen Fluoride: This process w a s studied i n d e t a i l by J. A. Donohue e t al at American O i l Co. (C-6,8) under contract DA- 31-124-A~0( D)-78. Original discovery of t h e method was made by Engelbrecht and Nachbauer i n 1959 (C-10).

4. Direct Electrolysis of Oxygen i n hydrogen f luor ide using a Porous Anode: This method recently patented by W. V. Childs at Ph i l l i p s Petroleum (C-5) involves passing a i r o r oxypen in to a porous carbon anode where it reacts with f luorine t o liberate OFZ.

The fluorine-caustic process was chosen fo r t h e i n i t i a l experimental work f o r several reasons; the process i s t h e most exten- s ive ly investigated and required the least investment and development cost of t h e four considered. Experiments performed with the fluorine-caustic system at fluorine rates up t o 2 l h / h r show no reduction i n y ie lds from l i t e r a t u r e values. Some experiments have a lso been done on the porous anode process.

5

111. Literature Review

Methods f o r making OF2 can be divided i n t o three main c lasses . These include (1) react ions of fluorine with inorganic compounds ( 2 ) react ions of f luor ine and oxygen and ( 3 ) d i rec t e l ec t ro lys i s .

A. Reactions of Fluorine with Inorganics

A-1. Fluorine-Caustic

Histor ical ly OF2 w a s first observed i n t h e of f gas of f luor ine c e l l s containing small amounts of water i n the e l ec t ro ly t e by Lebeau and Damiens (C-13) . Soon a f t e r t h e i r discovery they were able t o prepare OF2 by react ing f luorine w i t h 2% caustic ( A - 9 ) . A year later Ruff and Menzel (A-13) prepared OF2 by t h e caus t ic method and fur ther characterized it. During the next f i f t e e n years several authors reported y ie lds of OF using var ia t ions of t h e caustic-fluorine process (A- , 5, 8, 15, 16, 1 9 ) . f luorine involved bubbling excess f luor ine through a t h i n layer of 2% caust ic . This technique w a s used so t h a t the competing oxygen producing react ion would be minimized. The two reactions of f luor ine and caus t ic a re :

Early reactions w i t h caustic-

2F2 + 2NaOH->2NaF + OF2 + H20

2F2 + kNaOH,->bNaF + 02 + 2H20

I n t h e ear ly 1960's Harshaw ( A - 1 8 ) and Thiokol (A-7) prepared OF2 i n laboratory quant i t ies by contacting f luor ine with a t h i n f i lm of 2% NaOH i n wetted w a l l columns. I n both cases an excess of f luor ine w a s used. The Harshaw workers obtained a 70% yie ld . at Thiokol were only able t o obtain a 455 y ie ld using a 100 cm column giving a 25.5 sec. residence t i m e . decreasing t h e column length t o 50 cm workers at "hiokol obtained y ie lds over 70% at a 10.5 sec. residence t i m e .

Experimenters

By

I n a recent patent of Brown (A-21, y ie lds up t o 85% are claimed using l a rge excesses of caust ic . H i s claim is t h a t t he main source of oxygen is r ea l ly the decomposi- t i o n of OF2.

OF2 + 2NaOHr->2NaF + H20 + 02

6

This decomposition react ion is accelarated by the heat generated i n OF2 reaction. and keeping t h e temperature between 0 t o 25OC, he has obtained y ie lds between 70 and 85%. literature results are summarized i n Table I.

By using excesses of caus t ic

The f luor ine caust ic

Ruff and Kwasnik (A-13) prepared OF2 i n a two s t ep fluorine-caustic react ion with NO3F as an intermediate by t h e following react ion sequence.

2N03F + 2 N a O H 4 2 N d O 3 + OF2 + H20

Yields f o r t h i s react ion were not reported. do mention t h a t explosions w i t h N03F were common i f t h e proper react ion conditions f o r the n i t r i c ac id react ion were not maintained.

The authors

A-2. Fluorine-Water Reactions

Several other react ions of f luorine with inorganic compounds have been reported. The basis f o r several of these i s t h e react ion of fluorine and water.

2F2 + H20 4 2 H P + OF2

This has been s tudied i n de t a i l by R u f f (A-15) and Cady (A-5). Although both reported obtaining quan t i t i e s of OF2, y ie lds were very low with l a rge amounts of oxygen being formed. decomposition of OF2. f luor ine and ice-water (A-5). periodic acid (A-12) and so l id hydroxides of sodium, barium, potassium and calcium (A-5) . ac id react ion is ac tua l ly t h e water f luor ine react ion.

The oxygen w a s probably formed by t h e Cady reported a 5% y ie ld using

Cady a lso made OF2 from

The periodic

4F2 + HI04 2 H 2 0 4 2 0 F 2 + HI04

2F2 + HI04 2H20-02 + 4HF + H I 0 4

OF2 accounted f o r 27% of t h e gaseous products of t he reaction.

Rohrbach and Cady reported (A-11) OF2 and 02 i n the o f f gas from the reect ion of 60% perchloric acid w i t h fluo- r ine. Oxygen dif luoride comprised 45 mole % o f t h e of f gas. The OF2 and O2 were probably formed f r o m the water- f luoride reaction. The reaction desired w a s

7

4 rl 0 0

aJ .I4 x m

tQ d 0 rl

h 5: 0 a! 2

M cu k 0 k

0

€J 0

cu 0 c,

B 0 O S dc,

I b o O C 0 cu

0 c, 0 rl

$ 3 r l r l 0 0 0 0

ln

0 r i 8 al 2

c 0 k c +J

Y c 0 k .c 4J

Y c Y R c c,

r lc

t t % 0

c,

R cu a

al rl P P ? P

rl

v\

a aJ

..

3

a 0) rl P P s F9

a R c u i m cu

R

a a rl rl aJ aJ a .d .I4

rl x x 0, .d M M

0 ln a3

bp I I 0 In 0 t- 3 t-

GI h 0

a rl aJ

.rl x GI o m

al = b o 0 t-

w 0 UI

ln

0 0

rl I I

I m I cu cu cu cu cu

m k s 5:

a s 5:

2 s 8

al 2

x

a 2

0 rl

cr) I

rl

u) I cu I

I I I

cu rl

cu f

* m 0 w -

c3 a 4

n ln

32 V

h

cu

0- $ 4

pa" N 4 rl - 0 X 8

F~ + HCIO4,HF + ~ ~ 1 0 4 .

The react ion of a 31% solut ion H202 with f luor ine d i lu t ed with nitrogen has a l s o been reported (A-7). The react ion was conducted a t -18Oc with nitrogen t o prevent explosions and fires. Although some OF2 w a s obtained no y ie lds were reported and it i s uncertain whether H202 o r water react- ing i . i th t h e f luor ine ac tua l ly gave OF2. A t t he y i e lds given i n t h e l i t e r a t u r e these reactions can not compete with the fluorine-caustic process.

The most recent f luor ine water react ion i s t h e catalyzed react ion using a a l k a l i f luoride ca t a lys t . This process w a s disclosed i n 1969 by Borning and Pullen (A-3).

Barning and Pullen passed f luorine i n t o a glass vial containing an a l k a l i f luor ide and water. The water was condensed on the surface of CsF or NaF anhydrous power or added as a hydrate when potassium f luor ide w a s used as t h e a l k a l i f luoride.

The y i e ld f o r t h e react ion varied between 55 and 80%. The authors claim b e t t e r yields with KF02H20 and KF (TO%+) than w i t h NaF (60%). t o lump during the react ion and ag i t a t ion of t he alkali f luor ide was necessary.

The f luor ide mixture tended

Mr. Pullen was contacted regarding the experiments. H e stated t h a t no addi t ional work had been done and t h a t no o ther reports were available. The work had been conducted i n a few weeks and tha t more things would have been done i f t i m e permitted. H e thought t h a t by d i lu t ing t h e fluo- r i n e with nitrogen so as t o decrease t h e heat of react ion t h a t better y ie lds could be obtained. H e a l so stated t h a t even though their experiments showed b e t t e r y i e lds with KF than NaF, t h a t NaF would be b e t t e r on a l a rge r s ca l e react ion because of cost , ease of regeneration, and lessening of t h e caking problem. t o HF s ince none occurred until the react ion s t a r t ed .

H e a t t r i bu ted the caking

I n general t h e process appears t o o f f e r no concrete advantage over the fluorine-caustic method. It would probably require a more expensive reac tor , plus t h e regeneration of a l k a l i f luor ide would be an addi t ional processing step.

9

A-3. Other Fluorine-Inorganic Reactions

B r i e f mention has been found of t h ree other reactions of t h i s type of reaction although no y ie lds have been obtained. of t h e cesium f luoride catalyzed react ion of f luorine and potassium carbonate (A-10) and from t h e react ion of f luorine and potassium chlorate (A-17). Chlorine pentafluoride as the fluorinating agent has been reported t o react w i t h ozone t o give CIF3 and OF2 (A-1). These reactions are regarded as laboratory curiosi- ties, and w i l l not be considered fur ther .

Oxygen difluoride was found as a by-product

B. Direct Combination of Oxygen and Fluorine

Combination of oxygen and fluorine w a s first performed by huff and Menzel (B-10, Ell). By using an e l e c t r i c arc a t l i q u i d a i r temperature Ruff and Menzel w e r e able t o make 02F2. phase at low pressures. 02 t o F2, t h e current and voltage O3F2 (B-1, 2, 7, 8, 12) and O4F2 (€3-4) were prepared using t h e same basic low temperature and low pressure technique. and Grosse (B-13) later prepared owgen f luorides by t h e same technology i n t h e l iqu id phase. No details o r y ie lds are given and which oxygen f luorides were obtained i s not reported.

The reaction was performed i n the gas Later by varying the r a t i o of

Streng

I n 1966 Goetschel (B-3) obtained m a s s sp re t r a data f o r OF2 i n t h e vapors d i s t i l l i n g off a mixture assumed t o be higher order oxygen fluoride. "his material w a s prepared by low temperature i r rad ia t ion of t h e elements. The previously reported decomposition series f o r oxygen f luoride had not included OF2.

I n 1969 Kirshenbaum (B-6) reported a 20% conversion of OF2 from u l t r av io l e t i r rad ia t ion of ozone and f luorine at 120°K using approximately a 2:l ozone t o f luorine r a t io . 02 and F2. at 1 9 5 O K resu l ted i n only a 3 t o 51 conversion t o OF2.

The remaining gas included 02F2 (36% conversion) An equimolar mixture of 03 and F2 i r rad ia ted

10

C.

No report has been found where O2 and F2 were combined without i r r ad ia t ion or d i rec t e l ec t ro lys i s . A Pennsalt report states t h a t O2 and F2 reacted f o r 16 hours a t 375OC and 2100 psig yielded only F2 and O2 (B-9).

The lack of any pos i t ive results f o r OF2 production from oxygen-fluorine react ions preclude any economic calcu- l a t ions f o r t h i s type. The work Goetschel (B-3) indi- cates t h a t it may be possible t o prepare OF2 by low temperature i r r ad ia t ions but de f in i t e y ie lds i n excess of 35% w i l l be needed before it can compete with the f luor ine caus t ic published yield of 70% i n r a w mater ia l p r i ce alone. Since the reactions a re car r ied out at low temperature Kith W and/or high voltage equipment, a higher c a p i t a l investment i s inevi table . Therefore, fu r the r y i e ld improvements would be necessary i n order t o make fluorine-oxygen reactions p rac t i ca l .

The work of Kirshenbaum (B-6) with ozone and f luor ine represents t h e only quant i ta t ive r e s u l t s f o r t h i s type of reaction. converted with a 20% conversion t o OF2. I f t h e fluo- r i n e could be recycled the yield would be equivalent t o the base case 70% y ie ld for t h e fluorine-caustic process (s ince all the f luorine goes t o product only one ha l f t h e y i e ld i s necessary). Goetschel reports no y ie lds f o r any of h i s oxygen f luor ide work.

H e found t h a t 56% of t h e f luor ine w a s

Direct Elec t ro lys i s

C-1. Water i n HF

Although OF2 w a s f i r s t discovered i n t he anode gas of a f luor ine c e l l by Lebeau and Damiens ( C - 1 3 1 , no work w a s done t o optimize t h e production of OF2 by electro- l y s i s of w e t HF u n t i l t he last t e n years. Several authors discussed t h e poss ib i l i t y of producing OF2 e l e c t r o l y t i c a l l y (C-11, 16) but no e-erimental y ie lds were reported u n t i l Engelbrecht and Nachbauer published t h e i r work i n 1959 (C-10). They reported OF2 yie lds of 55 t o 60% over a water concentration range of 1 t o 20% using 10% NaF as a conductivity addi t ive. y ie lds were reported as % OF2 i n t he anode gas Kith the only impurity given as oxygen.

The OF2

Since both OF2 and

11

Water Content, mole % Metal Alkali , mole 7 In te r rupt ing current Current density, Amp/ft2 Maximum y i e l d (current eff . ) , % Water content 8 max. y i e ld , mole Fluor i ne p re 8 ent Temperature, OC

Voltage, Volts

0.1 t o 1.0 Yes 10-20 45

% 0.56 None 10-15 5-7.5

1.0 0.1 t o 1.0 0.2 t o 0.5 No mention 50-70 64 0.26 10% Max. -30 t o +50 6.4-8.3

The patent makes no mention o f t h e in te r rupt ing current feature stressed i n t h e contract report . gives higher y ie lds by operating near t h e threshold voltage of 8 vo l t s f o r f luorine production. between 7.0 and 8.3 vo l t s Donohue reports i n h i s patent up t o 10% f luor ine i n t h e effluent gas. a l s o mentions tha t a w a t e r content less than 0.2% gave severe anode corrosion.

The patent also

A t voltages

The contract report

olcygen require four Faradeys per mole, t h e e l e c t r i c a l y i e l d would be the same as the % OF2 i n the off gas at 100% anode eff ic iency. Engelbrecht, however, does not mention what anode efficiency w a s obtained.

A de t a i l ed study of the e lec t ro lys i s of wet HF was conducted at Americal O i l Company under contract DA- 31-124-ARO ( D ) - 78. contract reports , patents , and journal a r t i c l e s ( C-2, 3, 4, 6, 7, 8) by J. A. Donohue e t al. ported t h a t m a x i m u m y ie lds (current e f f i c i e n c i e s ) of 45% could be obtained over a much narrower range than t h a t reported by Engelbrecht. Donohue found t h a t maximum y i e lds could be obtained with water concen- t r a t i o n s of from 0.2 t o 2.0% mole and with a l k a l i f luor ides present i n about 0.2 t o 0.5 mole %. By increasing t h e water content ozone began forming i n preference t o OF2 up t o about 6% water content when t h e ozone production f e l l off. I n t h i s water range the OF2 y ie ld w a s about 10-15% and the current density f e l l o f f . Donohue suggests tha t above 8% water and 2% alkali f luor ide the main product would be oxygen although no data are given i n t h e 8 t o 20% water range and at 10% a l k a l i f luor ide metal where Engelbrecht re- por t s high y ie lds .

The work w a s reported i n several

Donohue re-

Donohue's f i n a l contract report (c-6) and his OF2 patent (C-8) present somewhat contrasting data. These axe compared i n the following t ab le :

Contract Reports U.S. Pat. 3,276,981

12

C-2. Water and Organic Compounds

I n h i s ear ly work on electrof luorinat ion of organic compounds, J. H. Simons reported OF2 present i n t h e off gases when water was added f o r conductivity o r i f oxygen containing compounds (eg organic ac ids ) were electrof luorinated (C-19) . I n 1966 Kisaki et al (C-12) i n t ry ing t o f ind which type of compound gave the best perfluoroalkyl carbonyl f luor ide (RfCOF) y i e ld a l so presented y ie ld data f o r OF2 and COF2. The best OF2 y ie lds are 20% for acids and 35% f o r aldehydes. Kisaki data f o r acet ic and Simons f o r butyr ic both give 20% y ie lds for OF2.

The e l ec t ro lys i s of ace t ic acid i s shown by t h e fo l - lowing reaction:

~ H F + CH3 COOH+CF3COF + OF2 + 5H2 (C-12)

The 10 Faradays needed t o complete t h i s react ion make it more expensive (operating and c a p i t a l investment) than t h e e l ec t ro lys i s of water which requires only 4 Faradays per mole unless much higher y i e l d could be obtained. Since only 6 Faradays would be required per mole f o r formic acid it would be b e t t e r than ace- t i c i f t h e same y ie lds could be obtained.

K i s a k i showed a 35% y i e l d of OF2 from acetaldehyde. However, Nagase (C-14) found a great deal of polymeric residue when electrof luorinat ing aldehydes. caused by t h e decomposition of f luor ina ted aldehydes i n the c e l l . shorts which would result i n increased maintenance costs. s i r a b l e than acids.

This w a s

Polymers i n t h e c e l l would produce

For these reasons aldehydes would be less de-

Nagase e t a1 (C-15) added water t o e thyl and propyl esters and alcohols. H e concluded t h a t s ince t h e amount of e l e c t r i c i t y needed t o produce any perflu- or inated compound varied d i rec t ly with t h e water added t h a t water w a s e lectrolyzed preferen t ia l ly over the organic alcohols and e s t e r s .

I

13

C-3. Direct Elec t ro lys i s of Inorganic Compounds

Oxygen d i f luor ide has been reported as an anode product of inorganic compounds by Rogers (c-16, 17) and Engelbrecht (C-9). t he major anode products of the e l e c t r o l y s i s of nitrogen oxides. of OF2 was reported with large amounts of current going anode losses .

Rogers reported OF2 and NF3 t o be

When y ie lds were calculated only a 2% y i e l d

Engelbrecht e lectrof luorinated a number of sulfur, o q g e n and nitrogen containing inorganics. and ammonium sulfates and persulfates yielded OF2, S02F2, N2, 02 and various other compounds including C02 from t h e polyvinyl chloride diaphragm. were common i n t h e piping of the divided c e l l . explosions were a t t r i bu ted t o NO3F. OF2 were obtained f o r potassium and ammonium persu l fa te . Early i n one K2S208 run the percent OF2 i n t h e anode gas ran as high as 60%. Later i n the same experiment com- parable amounts of S02F2 were obtained. From t h e data presented it w a s impossible t o ca lcu la te t h e amount of charge consumed during OF2 formation. tha t OF2 w a s formed mainly from water present i n the HF since using KF*2HF as the e l ec t ro ly t e at a higher temperature gave i n i t i a l l y only 12% OF2 i n the anode gas.

Potassium

Explosions Some

The bes t y i e lds of

It is probable

C-4. Direct Elec t ro lys i s Using Oxygen

I n a very recent patent (C-5) Childs at P h i l l i p s Petro- leum has claimed tha t it is poss ib le t o produce OF2 at 78% current eff ic iency by adding oxygen through a porous anode. He claims that oxygen snd f luor ine reac t i n the pores of t he electrode t o form OF2. H e states t h a t e lectrodes may be made from porous m e t a l s (eg. nickel , s t e e l ) o r w i t h porous carbon which are desired. advantages over t he water route i f moderate current e f f i c i enc ie s can be maintained.

This route t o OF2 appears t o have economic

t The process involves adding oxygen o r air i n t o a porous carbon anode. As t he oqgen passes through t h e anode it reac t s w i t h f luor ine being liberated from HF. The c e l l w i t h the exception of the anode construction and s m a l l e l ec t ro ly t e modifications is e s s e n t i a l l y a typ ica l f luorine c e l l .

The use of a porous anode for producing OF2 has been reported i n t h e l i t e r a t u r e . t h a t OF2 w a s produced by passing water through a porous carbon anode i n a typ ica l medium temperature f luor ine c e l l . No experimental data or y ie lds were presented. Donohue (C-2) tr ied t o introduce oxygen i n t o h i s elec- t r o l y t i c c e l l containing HF through a porous nickel anode. He reported t h a t when t h e addition was attempted tha t the pores plugged and no oxygen passed i n t o t h e c e l l . cathode and was able t o regenerate water necessary fo r t he e l ec t ro lys i s .

Allen (C-1) mentioned

H e then added oxygen through a porous nickel

The l i t e r a t u r e data for the d i r ec t e l e c t r o l y t i c methods a re compared i n Table 11.

1 5

I I

1 a" M

i k 0

B z c u c -0 cv

4J o m c,Q,

k m

k z 0 ln

C.

P- cu o\ rl

- m

Q,u a- s 7 s

rl rl aJ 0

a d a"

s

P

$ .)

m c, rl

co I

v)

.)

u 0 ln rl I 0

a rl aJ

.I4 h

I V . OF:, Production Economics

A-1. Basic Assumptions

Yields

"he methods of OF2 production considered i n t h i s report have been divided i n t o three classes. are (1) fluorine reactions with inorganics reactions with owgen and These processes all involve d i rec t ly o r ind i rec t ly an e l ec t ro lys i s reaction since the only commercial source of f luorine i s e lec t ro ly t ic . For f luorine reactions with inorganics and d i r ec t e lec t ro lys i s 100% yie ld would result i n 1.11 l b OF2 produced per 1000 Amp-hr of elec- t r i c i t y as shown by the following equations.

These processes (2) f luorine

(3) direct e lec t ro lys i s .

Fluorine production:

2F2 + 2YOH--J2YF + OF2 + H20

where Y i s usually H, K, o r N a

4 Faradays of e l e c t r i c i t y produce 1 mole OF2

Direct Electrolysis : (H20 as an example)

HF 2 H2 + 4F

4F + H20 -2 HF + OF2

4 Faradays of e l e c t r i c i t y produce 1 mole OF2

= 1.11 x 10-3 lb/Amp-hr (:::)(&::-)(3(-) o r 1.11 lb/1000 Amp-hr

Since operating costs and capi ta l investment are d i rec t ly proportional t o c e l l ampere rat ing, t h e lbs obtained per 1000 Amp-hr is the most meaningful way t o express yields f o r e l ec t ro ly t i c processes.

In t h e d i r ec t combination of oxygen and f luorine 100% y ie ld would give 2.22 l b of OF2 per 1000 Amp-hr.

4HF 4 Faradays 2*2 +

2 Faradays of e l e c t r i c i t y y i e l d 1 mole of OF2

Therefore the d i r ec t combination has an inherent advantage over t h e other two processes at equal y ie lds .

The e l e c t r o l y t i c method using oxygen i n a porous elec- t rode has the same advantages, s ince it i s t h e d i r ec t combination of oxygen and f luorine with a l l t h e f luor ine going t o OF2. This method then gives 2.22 l b of OF2 per 1000 Amp-hr at 100% yie ld .

B. Process Comparisons

Based on t h e literature review of OF2 production methods, four processes would seem plausible on a commercial scale . These processes are 1) fluorine-caustic 2 ) f luoride-water using an a l k a l i f luoride ca ta lys t l y s i s of water i n HF and 4 ) e lec t ro lys i s using oxygen i n a porous electrode. The advantages and disadvantages of these processes are given below.

3) electro-

B-1. Fluorine-Caustic

The fluorine-caustic is t h e oldest and best known process fo r preparing OF2. caus t ic process has been developed and operated a t fluo- r i n e rates up t o 2 lb /h r . output of a f luor ine c e l l would be 7 l b F2/hr. Further development cos ts t o get t h e process t o a commercial s i z e would be minimal which i s t h e biggest advantage of t h e fluorine-caustic process. t h i s process is t h a t t h e m i x t u r e of moisture, caus t ic and f luor ine is a very corrosive and reac t ive system. This r e s u l t s i n high maintenance costs f o r equipment, i n high labor costs and i n low on-stream ef f ic ienc ies . The react ion mixture can e e l o d e o r burn i f prec ise conditions are not maintained. For this reason labor costs may be higher than fo r t he other processes considered.

During t h i s contract t h e fluorine-

A commercial un i t using t h e

The main disadvantages of

18

B-2. Fluorine-Water (Alkali Fluoride Catalyzed)

This process i s based on a br ief a r t i c l e by Borning and Pullen using about 0.5 grams of f luor ine i n a 500 cc g lass bulb. There has been no sca le up of t h i s react ion toward a 7 lb F2/hr u n i t . Problems with t h i s react ion would probably be similar t o those i n the fluorine-caustic i n respect t o corrosion. The problem of caking of the reactants when HF is formed would have t o be solved. The react ion equipment, e i t h e r a corrosion res i s tan t s t i r r e d autoclave o r f lu id ized bed, would be more expensive than t h e fluorine-caustic react ion system. In summary, t he process o f f e r s no y i e ld o r other advantages over t h e fluorine-caustic system and would cost s ign i f i can t ly more t o develop t o a commercial un i t .

B-3. Elec t ro lys i s of Water i n HF

The process is based on the work of Donohue, done at American O i l under government contract DA-31-124-ARO (D)-78 and patented i n U.S. Patent 3,276,981. One possible advantage of t h e process i s t h a t t h e handling of f luor ine m a y be avoided. A second advantage i s t h a t t h e amount of HF used f o r e l e c t r o l y t i c processes is less than t h a t f o r t he fluorine react ion processes. The disadvantages are: 1) the process has not been scaled up pas t a 10 amp c e l l . A commercial c e l l f o r OF2 production by t h i s method would probably be between 4,000 and 6,000 amps. i n government reports a re below those of t h e f luor ine process, s ince 100% yie lds f o r the f luor ine processes and t h e e l ec t ro lys i s of water tire both 1.11 l b OF2/1000 amp-hr. Yields reported for the e l e c t r o l y s i s of water are 45 t o 64% compared t o 70% fo r f luor ine processes. This means t h a t t he labor costs o r t he investment costs w i l l be grea te r unless a greater on-stream fac to r can be obtained.

2) The yie lds given

B-4. E lec t ro lys i s Using Oxygen i n HF with a Porous Anode

The biggest advantage is t h a t t h i s process at 100% y i e l d gives 2.22 l b OF2/1000 amp-hr. s t a n t i a l l y reduced labor cos ts i n equivalent rated c e l l s at equal y ie lds . It a l s o has the other advantage, l i s ted above f o r e l ec t ro ly t i c processes. The y i e l d reported i s also as good as t h e f luor ine processes. The disadvantage i s t h e process a l s o has

This means sub-

not been scaled up. (U.S. Patent 3,461,049) w a s done with a 3 amp c e l l with a 30 cm surface area anode. Again a commercial c e l l would operate at 4,000 t o 6,000 amps.

The work reported i n t h e patent

2

C. Factors Affecting the Se l l ing Price

C-1. Design Criteria

"he design c r i t e r i a by which 10,000 lb/yr . could be made by each process are l i s t e d i n Table 111. The e l e c t r i c a l eff ic iency of each c e l l i s assumed t o be 90% (1.4 l b F2/1000 Amp-hr. or 2.00 l b OF2/1000 Amp- hr. for porous anode). weeks/yr i s assumed t o be 40. f o r process maintenance are included i n t h e labor man-hour figures. cesses considered w a s 66%. f luorine t h i s f igure would appear t o be r e a l i s t i c and perhaps higher than what could be real ized i n the i n i t i a l stages of operation. be able t o operate at o r above t h e 66% on-stream fac tor used.

The maximum number of operating An addi t ional 12 weeks

"he on-stream fac to r fo r all pro- For processes involving

The e l ec t ro ly t i c processes should

The t ab le a l s o l is ts t h e labor requirements used i n t h e calculat ions. The estimates for t h e fluorine-caustic process axe based on the experiments car r ied out during t h i s contract . "he operator man-hours fo r t he other f luorine process should be the same. Elec t ro ly t ic processes should require less manpower since the OF2 reactor has been eliminated. One less man/day f o r t he electro- l y t i c process w a s assumed although further work on sca le up of t h e process may indicate t h a t even less manpower i s needed. I n any case, th ree men/day (1 per s h i f t ) would be a dnimum uni t .

C-2. Investment Factors

The estimated investments f o r 10,000 and 20,000 lb /yr plants are given i n Tables I V and V. The 10,000 lb /y r plant would be used par t time for a 4,000 lb /yr commitment. A t production rates above 10,000 lb /y r t h e plant would have t o be operated 7 days/wk f o r ra tes up t o about 14,000 lb /hr , or addi t ional investment f o r more f luorine capacity and larger production and pur i f ica t ion equipment would be needed. The point where additional investment i s necessary would create a discontinuity i n the s e l l i n g p r i ce curve given i n Figure I.

A schematic breakdown of the OF2 processes is shown i n Figure 111. The figure shows e ight components which make up a typ ica l OF2 process. These eight components also represent e ight areas of capi ta l investment. These areas are plant site, f luorine generation, OF2 generation, OF2

20

H H H

aJ rl

R ti

rn aJ rn ul aJ 0 0 n" 8 3

cu

0 z P k e

x B

al 1-I k aJ

k u

d rn al F3

0 0

0 ri

O m

0 0 0 0 ri

m

0 0 0 0 Pi

m

ri cu ri kl

0

. . on L-

d cu d aJ v o 8 NOmWLn ; t \o

O W OCO

R TI o m cu

d cu V O W k 0 0 0 - 0 f c u

c u m ri R n L-

g 0 R

2 0,- 0 =tcu

ri I+ a 0 V O U k 0 0

cu d Frtm t-

0 o o m t -mwt -

Ln o o l n . mo\-.fcu

00 cum r! Lnd

00 OCO

00 Ln OCO

O N d

*m O m o o m c - m u m

21

k 0 0 QtO 0

m m 4JO

t 4 a +, ra

E H

0 0 O m In c-

0 0

0 \D

0,

00 00

00 00 O m "7,

0 0

m cu

0 0

P- rl

"7,

0 0

cu rl

%

0 0

cu t-i

%

-e3

m

E! v-4 4J al 0 4 Qt rl

k cu k 0

0 0

In cu O m

0 0

m ti

cum

0 0

m rl

0 0

o\ l-i

cum

e+ m

.rl

0

m m

ii 0 u

R 0 cu

0 0

f M rl

wm

0 0 In

m

$ rl

0 0

f 2 rl

-?!

0 0

f rl rl

-?m

* m

2 3 H

2 XI R 4J

0 0 0 0 4)

m

0 0 0 L n \D

n

0 0

0 In

0,

0 0 0 0 rl

.)

e3 m

4J

a 8 P e: 4

m ra Q) 0

l 22

n

0 W

m 0 0 0,

0 0 0 0 O m -e

0 0

in cu Oc

0 0

0 cu Oc

0 0

0 cu Oc

0 0

0 cu 0,

0 0

f m rl

-e

0 0

rl a rl

O m

0 0

rl In cu

"0,

0 0

cu 0 cu

*e

0 0

0 a3

O m

0 0

In \o

Oc

0 0

0 In

Oc

0 0

0 rl

Om

0 0 0 Pl a e

0 0

0 a 0,

0 0

Pl a 0,

0 0

m co 0,

i n m P - c u P-

in

0 0

0 rl rl

O m

0 0

rl 0

Oc

0 0

In cu "Om

0 0 a in cu c

0 0 L", 0 \D

n . k U

k Q) u + cu

k

00 00

0- 0a 0, Oe

0 0 *e cu P-

rl

0 d

2 u I cu

k

00 00 0-

0 0 cue c c

O W O m rl

+ m c

+e + c 8

d m g 0 u

z H

m

E H

bF,

Q)

cn

e

x k

c 8 rdd 8 u rl

rl A u

cu 3

23

rn Q) rn m Ql

cu E4 0

24

pur i f ica t ion , OF2 compression , OF2 cylinder investment OF2 storage and shipping inventory and process development. The storage and shipping inventory investment includes the working cap i t a l investment f o r OF2 product i n storage and shipment. t h e estimated investment necessary i n each area f o r the four processes considered.

Tables I V and V give

Both f luorine processes assume t h a t t h e OF2 is produced by f luorine generated from one commercial 5,000 amp f luorine c e l l . Since the cost of t h e f luor ine c e l l i s about half of t h e t o t a l investment, the most economical way t o produce 10,000 l b OF2 or less would be t o run t h e cel l a t full capacity f o r as long as necessary. Using a smaller c e l l i f only 4,000 lb/yr o r less are desired would be unwise since the saving i n using t h e labor only pa r t of 8 year would be a much greater saving than the f e w dol la rs of investment t h a t could be saved using a smaller cell .

Investment costs fo r the fluorine-caustic system should be r e a l i s t i c i f no additional equipment i s needed f o r a 7 l b F2/hr plant over t h a t required f o r t h e 2 l b F2/hr plant . f o r a l l the plants are similar, these values should a l l be qui te good. The electrolyzer investments especially f o r t h e porous anode process contain more uncertainty with the reactor investment f o r the fluorine-water process being the most uncertain. However, t h e i n i t i a l reactor and associated equipment cos ts f o r the fluorine- water system w i l l be greater than for t h e fluorine- caust ic system.

Since the pur i f ica t ion and compression equipment

The major pieces of equipment (eg. c e l l s and compressors) instrumentation and Diping should have a 1 0 year depre- c iable lifetime. Since the materials are corrosive and various pieces of equipment w i l l have t o be replaced, maintenance costs w i l l probably be high. In calculating t h e operating costs maintenance costs have been estimated at 10 and 8% of the fixed investments f o r t h e f luorine and e l ec t ro ly t i c processes respectively.

The other investment fac tor tha t weighs heavily on a process select ion i s the process development costs. Since t h e fluorine-caustic process is closest t o reduction t o prac t ice $10,000 w a s estimated fo r fu ture work. t h e other processes no scale up has been done past

For

25

small scale laboratory experiments. It would probably take a minimum from $50,000 t o $100,000 of development funds t o bring these processes t o t h e necessary s t a t e t o produce 10,000 l b OFZ/yr. The values given i n Tables I V and V are our best estimates of t h e funds required t o develop these processes. of t h e nuuibers r e f l e c t present thinking on t h e ease of developing the process and could vary as more ex- perience is gained i n developing the process. $80,000 value f o r process development of t h e porous anode e l ec t ro ly t i c process is t h e most uncertain of t h e development numbers given and i s probably somewhat opti- mistic. The fac t t h a t t h e development costs fo r t h i s process w i l l most l i ke ly be greater than any of t h e other processes considered i s re f lec ted i n t he values given i n Tables I V and V. Whatever t h e development costs are they would be included i n t he OF2 costs fo r an i n i t i a l OF2 production contract.

The comparative magnitude

The

C-3. Raw Materials

The pounds of various raw materials necessary per pound iE TohlP

V I . The fluorine-caustic is the most expensive process as far as r a w material costs are concerned; however, the raw material costs are only a minor pa r t of t h e operating cost fo r all t he processes.

=f nm- 0-n an+(-s+ad 4-1 +kr --~cI..c- . . . . h , . ~ m ~ - a e I- ------- --c -- -I------ --- ---- .----

C-4. Purif icat ion

A 90% pur i f ica t ion efficiency w a s assumed i n Table I11 as a result of pur i f ica t ion experiments reported i n t h e experimental section of t h i s report . efficiency includes collection as w e l l as purif icat ion losses. If no purif icat ion or co l lec t ion losses occur the se l l i ng pr ice could be reduced by a maximum of 10%.

This purif icat ion

D. Delivered Price

I n addition t o the se l l i ng price of OF2 three factors contribute t o the t o t a l delivered pr ice which must be paid f o r OF2. below are cylinder investment, ( l i qu id trailer cost i f capacity warrants ) ; shipping and storage inventory, which is a measure of the working cap i t a l needed; and t ransportat ion charges. from the s e l l i n g pr ice because they are independent of t he method used and depend on t h e details of a procurement.

These three factors described b r i e f ly

These charges have been omitted

26

a

m ch 0

m m 0 .

\D ch cu

In

0 'u.

I

0 In

0

c: 0

cu In

0 .

E- O

cu

cu cu 0

a3 cu cu

a3

I

I

0 . . P - \ D

rl

\o

I

I

cu d 0

cu d

0

cu \D

0

rl In

0

rl In

0 .

In 0 cu

3 0

Pi 0

0

rl 0 0

. . . . I . . . . c u O k + r l m I n 0

#-I

a Q) bo

4 &? 0

pc

8 bo

R +, rl P;

a

rl 4

8 ; 0 a rl

rl 2 4

a" 3 .rl

a rl

2 0 rl k 0 d

a"

e 4 k 0 rl

27

The ef fec t of these factors on t h e delivered p r i ce i s shown f o r t h e four processes for a 12 month contract i n Table V I I . The delivered price for a depreciated plant with no development cost for t h e fluorine-caustic and e l ec t ro ly t i c porous anode process i s shown i n Table V I I I . One possible cost estimate f o r each of these th ree factors i s described below.

D-1. Cylinder Investment

The OF2 could be stored and shipped as a gas i n cylinder .at 400 psig. 55-1/2" height) would contain 9 lb OF2. smaller cylinder a l s o used would hold 6-3/4 lb OF2.

New 9 l b and 6-3/4 l b cylinders with valves cost around $75/cylinder. are new o r have not been used previously fo r OF2 o r f luorine, t he cylinder must be cleaned, passivated, and evacuated before use. Procedures fo r cylinder preparation have been published ( D - 1 ) . "he cost fo r t h i s cylinder preparation would be minimal and would depend upon the condition of t h e cylinders.

The standard OF2 cylinder (10-5/8" O.D. x The s l igh t ly

I n addition i f the cylinders t o he used

ws an example me capi-cai inves-cmen-c i n GF2 cyiinciers would be $83,000 f o r t he 10,000 l b s of OF2 Droduction case (10,000 l b OF2/9 l b OF2 per cylinder x $75/cylinder = $83,000). t he investment write-off w o u l d be $41,500.

If the cylinders are depreciated t o half value

D-2. Shipping and Storage Inventory

The shipping and storage inventory is a measure of t he m d s spent fo r which no payment has been re- ceived usually cal led the working capi ta l . These values have been calculated and are presented i n Tables I V and V f o r t h e 10,000 and 20,000 l b OF2/hr assuming a 12 month period between allocation and payment. The a f fec t of these costs on the delivered pr ices are given i n Tables V I 1 and V I 1 1 for a s ing le 12 month contract and f o r a second contract.

D-3. Transportation Costs

Freight charges f o r shipping OF2 from U.S. east coast t o w e s t coast and returning t h e cylinder are $5/lb OF2 i n e i t h e r t he 9 or 6-3/4 lb cylinder. Shipment one way with no return i s $3/lb. f o r a s ing le 12 month contract and a second contract i n Tables V I 1 and V I I I .

These costs are shown

28

E m c u f r l t- l-i cu

0 000 0 Em 5 8 q 5 C O C O 0 0 v) t- O f l n c- cn r l c u M

0 000 0 sm:f e 0 t-00 t- a3 m 3 I n UI C O N cu

*I

m rl

m rl

s v) c u f L n t- cu m $

(d k +r c 0 u

comcuoo 8 8gzJ 0 8 m m m

8 Z881 8 N 0 0 0 (u c

+r

E 8 cu rl 0) l-i M c d

I

cn I n L n I n f Cu . InrloI . . . 0

I U

0 0 0 0 0

2Im sm UI C u r l 0 0 C O f f L n Cu m In

0 0 0 0 0 3001 2 0 cn In InLn k

0 ec al u

m o l n L n m I- O d O l m . . . . a

al k

f

rl al FI

0 000 0 m 0v)In 0 0 o o o l 0

a al

*I

"I rn

0 d

rl

+, 0 al d i al

0 d

&

0 cu 4

0000 0 o o o o l mC0Lnl.n $1 d *

I

k 0 +I 0 al 0 d

N

n" 0 m . m m m c o a In

- r l

a .d

C al

0 4 a k

0 0 f a

al * m P- m

iJ

rl a 9

m 0 0

PI k aJ 4J

m

E

4 aJ G d k

rl R 9

ri Pi k

Q, iJ

al 0 d k PI

al 0 d k PI a aJ SI

a al k aJ * d

aj * d

30

0 000 0 a 2 0 0 cu 0 o o o l 0 ;

", t a3 0 0 0 a3

m c u

\D r l L n l n L- v3 . ,do( . . . o Ln LnLnLn 0 cu c u r 4 0 1 c- 0 c u f l n cu cu I m

0 000 0 0 0 0 0 0 0 0001 0

rl c u l n l n m L- r l f i o l 0

H H E a, rl

2 0 0 0 0 0 2 a d 0 0 OOO] 0 0 0 0 g

m WLnLn m m m m a,

0 .rl k pc a E P d rl 9) n a Q)

d 3 4J 8

0 000 0 cu W W L n o\ 0 0001 0 ""Ln gg8J 0 8

L o w 0 f i r l c u l Ln In

+, a! 9) 0

O d k 9 &

.A I

31

E.

F.

G.

Production Lead Time

The t i m e needed t o construct and prove operat ional t h e 7 l b F2/hr f a c i l i t y fo r producing OF2 as described i n t h i s report would be between 9 and 12 months, depending on the prohlems encountered i n start-up. A t t h e end of t h i s period production of OF2 could be- gin using 7 l b of F*/hr.

I n f l a t i o n Factors

The s e l l i n g pr ices given i n Figures I and I1 would hold for contracts committed i n t h e f i s ca l year 1971. Pr ices beyond 1971 could possibly increase at a rate of 6 t o 8%/sr.

Multiple Year Contracts

"his report has been primarily concerned with t h e s e l l i n g p r i ce of OF2 produced by four d i f fe ren t processes f o r a s ing le 12 month contract . The change i n the s e l l i n g p r i ce f o r OF2 produced by the fluorine-caustic and e lec t - ro lys i s of owgen i n HF using a Dorous anode with no process development costs and depreciated equiDment was shown i n Figure 11. These reduced pr ices would apply t o a second contract after nlant and development costs have been met i n the f i r s t contract .

Figures IV and V give t h e cumulative s e l l i n g p r i ce f o r 10,000 t o 50,000 l b OF2 Droduced i n a one t o f ive year period f o r t h e fluorine-caustic and e l ec t ro ly t i c Dorous anode processes. The cumulative se l l ing pr ices were obtained using the costs given i n Figures I and 11. "he one year costs w e r e obtained from Figure I. For multiple year requirements it w a s assumed t h a t equal amounts would be produced each year with second contract Trices used for each year after t h e first. A minimum of 5,000 lbs w a s assumed produced each p a r . The figures show t h a t t h e t o t a l cost increases as t h e number of producing years increases. The main reason fo r t h i s increase i s t h a t t h e s ign i f icant labor costs per lb increase as t h e number of producing years increases. The f igures assume continuous prodaction and t h a t all t h e OF2 produced i n a year i s shipped and naid f o r within t h e year. i s deferred over several vears even though DrOdUCtiOn i s made i n one y e m the shipninq and storafre costs would bring t h e one year cost c loser t o multiple year costs .

If Dayment

32

FIGURE V

Cumulative Sell ing Price of OF2 at Various Production Levels by the Electrolysis of Oxygen i n HF for Various Contract Periods

Sell ing Price Does Not Include Transportation Working Capital and Storage Charges. Productions Assume Equal Production Each Year

M u l t i - Y e a r

34

v. Exp eriment al

A. Fluorine-Caustic System

A-1. Experiments at 0.1 and 0.2 lb F?/hr.

a. Experimental Procedures

A schematic representation of t h e experimental apparatus i s shown i n Figure V I . pictured i n Figures V I 1 and V I I I . ponents are the reactor , freezeout drier, C02 t r ap , and l i q u i f i e r .

The apparatus i s The major com-

The f luor ine i s fed from 8 cylinder t o a surge volume maintained at a constant pressure t o help maintain a constant flow rate t o the reactor. vessel it goes through a g l a s s flowmeter i n t o t h e reactor. make up tank through a flowmeter t o t h e reactor. The l iqu id level i n t h e reactor is maintained by a l i q u i d l e g i n t h e caust ic e x i t l i n e which i s vented t o prevent siphoning. The e x i t l i n e a lso contains a thermo- couple t o measure t h e temperature of the caust ic . The i Y u u L . i i i c eiiLers iiiruugii oririces w n i c n are usually four inches below t h e caustic surface. The oxygen and oxygen dif luoride formed e x i t through the top of t h e reactor.

From t h e surge

Dilute 2% caust ic solution i s pumped from a

Two reactors have been used. The first w a s a modi- f ied 500 cc gas washing: bot t le . reDlaced by a Elass bulb containing one, two o r four 0.1" o r i f i c e s through which t h e f luorine entered the system. be replaced frequently as the openings expand with use. A majority of t he runs except those with sparger depth var ia t ion have been done with t h e 2" diameter glass reactor. The second reactor was a 2 ft. lonp 2" diameter Fiece of t e f l o n pipe with machined caps on both ends. The i n l e t s and ou t l e t s w e r e t h e same as t h e glass reactor exceut f o r t h e f luor ine i n l e t which w a s a 1/4" piece of t e f lon pipe plugged at t h e bottom which can be raised o r lowered t o any depth. The bottom inch of the pipe w a s d r i l l e d with 70-0.020" holes. These reactors were used f o r t h e 0.1 (500 cc FZ/min) and 0.2 (1,000 cc F2/min) lb F2/hr runs.

The f r i t t e d disc w a s

These bulbs were severely attacked and had t c

35

I I ,I \ I

I 'I /

36

~

Figure V I I . OF2 Explosion Chamber

Figure VIII, OF2 0'1 l b F2/hr Reactor

37

The OF2 and oxygen ex i t i ng from the top of t h e reactor pass i n t o a 300 cc cylinder equipped with a dip tube. The cylinder i s placed i n a C02 cooled t r i c h l o r bath at -10O0F. The purpose o f t h e t r a p is t o freeze out any water enter ing w i t h t he product gases.

The second t r a p operated at about -150°F, removes C02 From the gas stream. The t r a p i s cooled by adding l i q u i d nitrogen t o R-11 o r methylene chloride u n t i l t h e l i qu id slushes. This t r a p w a s i n s t a l l e d after pers i s tan t plugging of l i q u i f i e r i n l e t shortened or wiped out several runs. From t h i s t r a p the product gases are l i q u i f i e d i n a 300 cc cylinder immersed i n l i qu id nitrogen.

The cold t raps a re evacuated at the beginning of t h e run and kept under vacuum during t h e run. The i n l e t l i n e t o the vacuum pump i s equipped w i t h a combination charcoal-soda l i m e trap plus a LN2 t r a p t o pro tec t t h e pump. scrubber packed w i t h charcoal, soda lime, K I c rys t a l s and rubber. Both scrubbers have worked well when used.

OF2 is vented f romthe receiver through a

A i L r r coiiruiion vi GF2 in iiie receiver for aboui 2 hours t h e receiver is warmed up and the pressure i s equalized with a 2150 cc evacuated cylinder. i s calculated from t h e OF2 analysis i n the cylinder and receiver , t he system pressure and the f luor ine reacted. Occasionally when the f luor ine flow deviates from the s e t point , t he y i e ld is corrected by using t h e system pressure which is a check on t h e f luor ine flow. The OF2 i n the cylinder o r receiver w a s analyzed by I R o r GC.

The y i e ld

b. Experimental D i f f i cu l t i e s

Corrosion - As i n a l l f luorine systems corrosion has been a problem i n t h e flowmeter, reactor and water freeze out t rap. have been generally f r e e of corrosion. and OF2 l i n e s a re copper, t raps are 304 S.S. or copper, and the valves are Hoke Monel valves, model number 343.

Constituents containing dry OF2 All f luor ine

30

Corrosion of t h e glass reactor i n l e t and flowmeter due t o HF were expected and have occurred. The i n l e t d ip tube on t h e freeze out t r a p plugged several t i m e s with green metal fluorides. able when m e t a l demister was placed i n t h e top of t h e reac tor at high f luorine flow rates t o prevent l i qu id carry over. The demister w a s removed as w a s m e t a l packing i n t h e freezeout t r a p which became corroded.

Plugging w a s most notice-

The f luorides a l so attacked the vanes i n the vacuum pump but l e f t t h e Kel-F o i l untouched. The problem was improved when t h e pump trapping system was in- stalled.

C02 - I n i t i a l runs at 500 cc/min f luorine gave l i t t l e or no evidence of C02 generation. Problems w i t h C02 plugs i n the receiver started when a new cylinder of f luorine w a s used. Although carbonates i n the water and caust ic can form C02 i n the presence of f luorine, C02 i n the f luorine was believed t o be the main source of t h e C02 problem. t i l l e d water t h e C02 problem halted t h e run before completion when a higher fluorine t o caust ic r a t i o was employed. since a s t i l l e d water sone c o u a not be counted on t o insure a C02 free system a C02 freezeout t r a p w a s ins ta l led . been observed using regular t a p water and caustic.

For t h e two runs done with dis-

With t h i s t r a p C02 plugs have

Operation - Although i n general operation of the experimental system was good once operational, t h e problem w a s get t ing a run started. Problems w i t h frozen l i n e s , plugs, leaks, and corrosion were prevelant on start up. This condition is common with f luorine systems and continuous operation i s desired i f possible.

Fluorine - Fluorine break through d id occur a t high f luorine t o caust ic r a t io s . This resul ted i n corrosion i n the water t r a p and created problems with the anal- y t i c a l equipment especially the AgCl windows on the i n f r a red analyzer. w a s less of a problem; however, some sens i t i v i ty was l o s t i n the instrument.

Fluorine in t h e gas chromatograph

39

Fires - One batch of caust ic w a s made up w i t h caust ic f'rom a la rge storage tank of 45% KOH. contained corrosion products or sediments from t h e tank and w a s d i r t y i n appearance. fires occurred at t h e f luorine i n l e t of both the g lass and t e f lon reactor. extinguished by shut t ing off t h e f luorine. In t h e t e f lon system black spots appeared on the w a l l s of t h e reactor and two holes were burned i n t h e sparger tube. and no further fires occurred in e i ther system.

This caust ic

When used numerous

The fires were quickly

A f t e r tha t fresh 45% caustic w a s purchased

c. o p erat ing Variables

The operating variables studied included f luor ine t o caust ic flow r a t i o , sparger depth, caust ic type, fluo- r i n e o r i f i c e veloci ty (bubble s ize) and agi ta t ion. Sample data i l l u s t r a t i n g the effect of these var ia t ions are given i n Table IX.

Most of t h e experiments were performed at a f luorine flow r a t e of 500 cc/min with caustic flow variat ions from 0.4 t o 1.3 based on the fluorine volumetric i n l e t flaw. The molar r a t i o of caustic t o f luorine f o r these flow rates are 3.4 t o 10.7. A t no t i m e w a s molar r a t i o of caust ic t o f luorine less than 1 t o 1. I n i t i a l experiments were performed using sodium hydroxide ; however, potassium hydroxide w a s used fo r most of t he runs because of t h e increased so lubi l i ty of potassium f luoride over sodium f luoride in an aqueous system.

Sparger posit ion and construction were a l s o studied. Experiments were performed w i t h t he sparger immersed i n four, eight and twelve inches of 2% caust ic ( R u n s 34-36). The o r i f i c e diameter varied from 0.02" t o 0.25". flow veloci ty through each or i f ice w a s maintained over a 10 t o 1 range by varying the number of o r i f i ce s .

The

Two runs were made w i t h a magnetic stirrer i n the bottom of the glass reactor (Run 37). Other runs were made wi th 5% potassium f luoride t o simulate the e f f ec t of recycling caustic. Since the caust ic w a s outside additional cooling w a s not required. few degrees higher than the i n l e t and averaged about 50°F.

The ef fec t of temperature w a s not studied.

The temperature o f t h e ex i t ing caust ic was a

40

0 0 In

0 0 In

0 0 In

0 Q In

0 0 0 rl

0 CR a

3 Ln m

0 cu rl

0 Cn

0 rl 4

W t- I In t-

3 t- I

a3 W

3 t- I cu

t-

o\ t- I a

W

m t-

I t- W

0 a3 I

3 W

3 UI

W UJ

W W

cu t-

cu t-

In W

m OD

m m

t- o\

0 0

orl t-\

ln t-

d@J * W

0 0 \ . N O

rlul * c u O d \ . a 0

r l r l . m 0 0 \ . 4 0

? Z e?

0 0 d

t- t- I

t- m

0 OD

cu 3

a3 t-

t- In

ln t-

d t- I t- W

o\ W

cu t-

a o\

I

0 cu rl

W t- I

3 W

cu t-

W o\

I r l . t- W

0 @J rl

t- t- I a3

W

W t- I a3 W

m t-

3 t-

d-o m m d

0 cu rl

(u Ln I cn

3

3 In I

t- 3

m In

I I

w mcu 03 94&i

o\ f I ln 3

m Ln

o m do\ rl\

I

t- Ln

I I

o t - 0 c u W W d I

0 W

t - 3 corn In o\ I I

I aJ \

h

. . . . m m - x x u

h

k 0 c, 0 ai aJ Iz 41

d. Discussion of Results

Several of t he main runs i l l u s t r a t i n g the effect of operating variables are given i n Table I X . Other experimental data are given i n Tables X, X I and XII. The ef fec t of various operating conditions are given below.

Type of Caustic

Most runs were done with KOH because of t h e increased so lub i l i t y of KF over t h a t of NaF. was noted but KOH experiments w e r e easier t o run because the NaF tended t o plug the f luorine sparger and caustic l i nes .

No y ie ld difference

Flow Ratio

Yields around 65% w e r e a t ta inable using a volumetric r a t i o of caust ic solution t o fluorine of a t least 1:l. A t a lower r a t i o t h e y ie ld decreased i n t o the 50's (Ftuns 12, 13, 18, 19) . A t a 0.5:l caust ic t o f luorine r a t i o at 1000 cc/min fluorine some breakthrough of f'liinFino nt-Tiirrerl at. Q h i n r h spraer i4ent.h-

Sparger Depth

Sparger depths of four, eight, and twelve inches were studied a t 1:l volumetric f low r a t io s . The y ie lds of a l l experiments were about 65%.

Sparger Construction

Orifice diameter (bubble s i ze ) had no noticeable a f fec t on the yields . The main purpose i n running with various s izes was t o study the corrosion on the glass and t e f lon sparger . Severe corrosion w a s noted with the glass sparger where a l iqu id pool of HF collected i n the bottom on the glass bulb and d i s so lved the bottom of t h e bulb. The glass o r i f i ce s were continually being enlarged by t he action of HF. The glass spargers l a s t ed two t o three runs and then had t o be replaced.

42

ri m o O f

21 1 cn cu ot-

- I I c u d a

zocu Ln o n x 3 Ln t-CO O f cu m n t - w c n w cu d ri

cu c o 0 . cu TI30 h \ Q O l 3 t -

0 u m o O f

dl cn ul

dt- IA

ot- I I Receiver Plugged zocu Ln

gz d

cu G O 0 . cu 4 3 0 h \ a J 0 l 3 t -

a3 0- * I I Nt-

Ln

x o r l ul OLn x f z

IA

0 t-w co Lnw ri

0 3 a 3 c n c n 3

cu G O 0 . cu r i 3 0 h \ Q O l 3 t -

f dI t- cu \ 0 rl

0 cn f

m O W

* I I cum Ln

ocut- m t - - m rl

in

m

b; f

0 cu \ 0 d

i cu dl x ocu

I 1 1 1 0 m t - d c n n d

t- xocu 0 0 I Ldcu 2

d \ 0 I l l k a 3 f

CO a m rl

D z o c u OCO I c d r i z

9 IJsed ?!aterial for Analytical 4 Calibrations

P;u H O

h h P P

r:

CJ 0

al L

cucu k R 0 0

cu M R

.rl 0

+r al P;

P) +r cd n

43

0 mt- 0 mt- c u \

W In

I n r l I t -k I

I I I

t- W

* 0 m m In m u r l \

0 W

N v ) ; t r l

0 \

9 rt

, I 3 I X

I E! * 0 m I n m m u

\ 0 W

cu f f l f r l

0 \ rl

bli cu 3

o m I I n k I

I I I

w

?3 5 R I u H €4 rn 3

I 4

m W

cv f f l * r l

0 \ rl

M cu 3

0 1 I r l I I

I I I I

0 4 0 c u d 0 No Yields

F2 Break Through

Back Pressure

O N In f

O N In I f N f f l f rl

0 \ rl

hi) d

$ 0 - I c u t - I rl I

I I d

f W

c o U f d \ I n

r l m rl

o t - W c u t - t- rl

I I

c o c o \ 0 \ 3

cu F4 0

c .vi E \ 0 0

aJ

*

44

P- m

W m

In m

f m

m m

cu m

rl m

0 m

co cu

G

W cu

4 In cu

3 a

\D \ rl

cu \ rl

0 (r7 \ cu rl

o\ cu \ cu 4

cu ti \ cu rl

rl rl \ cu rl

0 rl \ cu rl

o\ \ cu rl

In \ cu rl

W cu \ ti rl

In cu \ ti rl

f cu \ rl rl

al c1 aJ n

0 0 In

0 0 In

0 0 In

0 0 In

* 0 0 In

* 0 0 3

0 0 0 rl

0 0 In

0 0 In

0 0 In

0 z 0 0 0 rl

8

c -4 Ei \ 0 0

aJ +r aJ a cu b

x o Oo\NBB SCf u\

I

ca ca c u d

rl w

ri

0 \ cu

.

cv 0

\ 0 P-

cu 0 \ 0 P-

8

cu 0

\ 0 P-

0

d 0 \ cu

0

? 0 \ N

4

0 \ cu

rl

0 \ cu

t 0 \ f

? 0 \ f

t 0 \ f

d 0 \ f

1

k kaJ - .I,

45

0 0 rl

0 cu rl

0 o\

0 rl rl

v> f

0 o\

0 o\

0 cu rl

0 \D

0 3

In P-

O m

P- P- I m P-

W P- I In t-

f P- I co

W

f P-

P- 4

I

o\ P- I

W P-

I

I

I

I

I

I

0 co I cu

P-

o\ P- I t W

o\ P- I

P- W

0 co I f \o

In a

0 co I In t-

a3 P- I f P-

0 co I 0 t-

\D P- I f P-

E- \D I 0 \D

m W I 0 W

s

c u c u E L 0 0 = s a .

0 P-

In f ri

0 rl rl

In f rl

0 o\ \ 0 a0

0 P-

I

0 0 rl

I

I

I

I

0 0 r l c u \f in P-

o\ o\* \W co W

In

a * o\O \P- W W

0

In

P-* o \O \I- W W

cu 8 9

l r l 0 1 4 !z

I

4 k

I @ a aJco acu a " 5 -4

I Ulk

rl Pl

I c u 0 u

aJ k

u) Q) k a

2

s 2

z

1

rll H

c, 0 aJ d rl

cu 0 a

cu r=i 0 k 0

f rl

g c 0

a aJ m aJ P a rl aJ '4 x k 0 k %

g

0 '4

ca

rl 6.1

cu E4 *

Agitation

The y i e l d for t h e agi ta t ion run w a s only 42%. A second run gave a 98% yie ld based on f luorine added. The problem i s probably with t h e placement of t h e magnetic ag i ta tor i n t h e bottom of t h e reactor. The ag i ta tor draws fluorine o r OF2 down i n t o the reactor and out t h e caustic drain l i n e . Bubbles observed i n the l i n e confirm t h i s , but from t h e di- vergance of yields it would seem a s i f f luorine w a s removed i n one case and OF2 i n t h e other. From the on-stream analysis and product analysis both runs should have a 70+% yield. because of t he cylinder pressure. Product analysis for t he runs w a s typical . analysis r a t i o gave a 70% yield f o r both runs.

The y ie lds are diverted

Yields based on t h e 02/OF2

KF Addition

Runs with 5% added KF were similar t o other runs and gave about a 66% yield. If the caust ic is kept i n a closed system with no fresh caustic added, t h e y i e ld dropped off as the KOH concentration dropped t o one per cent.

General

In f r a red spectra and gas chromatograms were obtained on the f i n a l product and periodicly on the gas stream entering t h e receiver.

On-stream yie lds f o r runs using a 1:l flow r a t i o ranged from 70 t o 80% yet t h e f i n a l product yields calculated from t h e f luorine flow r a t e , cylinder pressure and OF2 analysis were always about 65%. If only t h e oxygen t o OF2 r a t i o is used, the yields average about 72%. This y ie ld assumes t h a t all the fluorine i s converted t o OF2 or 02 and none i s le f t unreacted. are made and no fluorine goes unreacted t h e pressure i n t h e receiver can be easi ly calculated. The r a t i o of t he actual and theore t ica l pressures are given i n Table I X . A low value would indicate that t h e f luorine rate i s o f f , leaks are present o r as i n the agi ta ted run t h e gas is being sucked out wi th the caustic. would indicate unreacted fluorine or vacuum side leaks.

If only 02 and OF2

A value over 100%

46

A-2. Experiments at 2.0 l b Fp/hr

a.

b.

C.

Procedures

A l a rger glass reactor made from the glass resin reactor w a s operated at fluorine rates up t o 3 l b F2/hr. The system w a s designed t o operate at 2 l b F2/hr. Two homemade copper t raps operated at -100 and -150'F were used fo r water and COP removal. The OF2 was l iquefied i n two 1000 m l high pressure S . S . cylinders. The rest of t h e system w a s essent ia l ly the same as before. The system was operated as the 0.1 l b F2/hr apparatus.

The major problem with the system revolved around the f luorine sparger. Teflon spargers were found unsatis- factory. The f luorine appeared t o burn t h e te f lon which led t o explosions in the reactor. These explosions were sometimes preceded by a clouding of t he caustic and followed by a black discharge from the sparger. ger was t e f lon but a l s o occurred at other t i m e s when no te f lon was present.

The explosions were prevalent when t h e spar-

Results

The results from these runs are summarized i n Table XIII. On-stream analysis have generally been over 70% OF2 and yields 65% or greater.

Purif icat ion

Some crude OF2 (76 mole % OF2, 23 mole % 02, 1 mole % C02) w a s pur i f ied by pusging with 0.5 t o 1 .0 SCFH of helium at 10 t o 20 mm Hg. every half hour f o r purity. A f t e r about 2 hours, t he OF2 was 98.6 mole %. between C02 and 02. and handling losses w a s over 80%. t h a t t h e C02 content did not decrease s ignif icant ly . The experiment proves t h a t OF2 and 02 can be separated without d i s t i l l a t i o n by purging with helium.

The product w a s checked

The remainder w a s equally divided The recovery with all the sampling

It should be noted

47

Table HI1

Ekperimental OF2 Data Using the 2 lb/hr Fluorine-Caustic System

RUn

Date

Rates - Fluorine , lb/hr

KOH, gpm

Reactor

Sparger Material

Fluorine Inlet Line

Duration, Minutes

On-stream Analysis, % OF2 by G.C.

Crude Product

receiver cy1 inder Pressure, psig

receiver Analysis, % OF2 cylinder

Yield, % (Based on F2 and pressure)

Yield, % (Based on 02/OF2 Ratio)

38

1/28

o 18-0.36 0.6

Teflon

Monel

110

72-73

175117 5

94/78

65 *

42

3/18

1.0

1.2

Teflon

Teflon

60

65-75

90/90

90 6/71

65

73

44

418

1.0

2.0

S.S.

Monel

120

79-81

2151195

97 5/79

73

87

45

419

3.0 2.0

S.S. Monel

25 75-77

1551135

90 5/75

91

82

*Vented Pressure During Run

48

B. The Electrolyt ic Oqygen Porous Anode System

B-1. Procedure

An experimental e l ec t ro ly t i c c e l l was set up as shown i n Figure I X . The c e l l w a s constructed from 4" monel pipe. The c e l l contained a central porous anode sur- rounded by a nickel diaphragm and a steel cathode. The volume of t h e c e l l w a s about 2 l i ters. Several designs for the porous anode have been t r i ed . The first anode w a s a 1-1/2" O.D. porous carbon cylinder plugged at both ends. This design w a s discarded f o r several reasons. None of t he sealants t r i e d could withstand the corrosion and e lec t ro ly te leaked i n t o the annulus and plugged t h e bottom of t h e gas i n l e t l i n e . The oqygen f o r OF2 i s added as a 25% mix with helium. The result w a s t h a t t h e oxygen e i the r leaked out the top or came through t h e pores a t t h e top of the anode. A second anode with a hole 3/4 of the way down t h e center w a s scrapped f o r similar reasons. The f i n a l anode tested introduced t h e oxygen i n t o the bottom of a 1" x 1" x 4" rectangular anode. The oxygen dis t r ibut ion w a s bet ter than the first two but s t i l l not what i s desired.

E%-2. Results

Several posi t ive results have come from these i n i t i a l experiments. There were no problems i n get t ing the c e l l t o conduct properly. f o r current densi t ies up t o 75 amps/ft2. duced with each anode design. evolved have been F2 and t races of C02 and COF2.

The voltage has been 5 t o 7 vol t s OF2 w a s pro-

The only other gases

The major problem has been i n obtaining desirable cur- ren t e f f ic ienc ies . The cwren t e f f ic ienc ies obtained are around 5% f o r these preliminary experiments. These values can be explained by poor oxygen d is t r ibu t ion i n t h e anode, oxygen leaks around the anode, and leaks i n t h e experimental apparatus. i n t he anode can be improved so t h a t f luorine manufacture is prevented improved current eff ic iencies should result.

I f the oxygen d is t r ibu t ion

49

[ OLL N W

I

I- 0 - cu - 0 0 0

0 I-

P

'I F z W > 0 I-

I cu

a 0 lL

3 a 0

k V W -J W

(z W

5

50

C. Analytical Procedure and Equipment

The OF2 produced has been analyzed by i n f r a red and gas chromatographic techniques. Using a 2 mm f ixed path gas, s i l v e r chloride windowed c e l l , a ca l ibra t ion curve w a s obtained f o r t h e f i r s t spike of t h e strong; t r i p l e t peak at 12 microns. The i n f r a red compared w e l l with chromatographic results. within 2 t o 3% of each other. results are used f o r calculations since they can be read more precisely especial ly a t OF2 values near 100%. Two d i f fe ren t gas chronatograph systems have been used. The first consisted of a Beckman GC-1 chromatograph operated a t room temperature with a 11 m l sample loop. The column w a s 4 ft. by 3/16" O.D. Y . S . tubing packed with Davison Grade 08 s i l i c a gel. worked w e l l but began t o lose sens i t iv i ty after f luorine analysis. The present instrument is a Beckman GC-2A chromatograph operated a t )+O°C. The column i s 10 ft. by 3/16" O.D. E . S . tubing packed w i t h Fisher Sc ien t i f i c Grade 28 s i l i c a gel. separates oxygen and OF2 with C02 t r a i l i n g out after 1 9 -:.....b..- A--. *l..,...,..., -- ..:+-̂-I - **.:I1 -1..,.3.. ..:+L

t h e oxygen. A Beckman integrator-recorder i s used t o record the results.

The two are usually The gas chromatograph

The instrument

The s i l i c a ge l used for both

-4 -.*-"I". '"'J A & u v I * . . - "I "*-""R-" ..-A* -......-- I.&..&.

I 51

V I . Conclusion

O f t he processes studied, t he fluorine-caustic process i s recommended for immediate production of OF2 on a commercial basis unless the quantity desired is la rge enough t o write off t h e process development costs associated with other processes (e. g. t h e oxygen porous anode process ) . The fluorine-caustic process i s suggested fo r t he following reasons: cess has been scaled up t o 2 l b F2/hr. compared t o t h e pro- jected production rate of 7 l b F2/hr. Other processes have only been operated on a gram scale 2 ) The scale-up has al- lowed more accurate estimation of the investment cos ts . The l a rges t investment factor i n t h e investment fo r t he fluorine- caust ic process i s the f luorine c e l l which has been designed and operated f o r many years. mJor piece of equipment has yet t o be designed and problems i n i t s operation are not known. been obtained on a laboratory and scaled up reactor f o r t he fluorine-caustic system. The main disadvantages of t he fluorine-caustic system are the low on-stream t i m e and high maintenance allowances.

1) The pro-

I n all the other processes a

3) Literature yields have

Although a l l the other possible OF2 processes have an operating cost advantage over t he fluorine-caustic system, the only one t h a t could be worth t h e gamble of scale-up i s the oxygen e l ec t ro ly t i c porous anode process because of t h e poten t ia l based on t h e superior l i terature yield. If these yields can be obtained comerc ia l ly the labor and investment costs would be s u b s t m t i a l l y below the fluorine-caustic method. I f t he t o t a l amount of OF2 t o be produced for NASA programs i s i n excess of 20,000 lb s it i s recommended t h a t some addi t ional development work should be done on the oxygen porous anode process i n a scaled up version of Childs 3 amp c e l l . If t h e y ie lds reported fo r t h e 3 amp c e l l can be r ea l i zed i n la rger commercial sized uni ts (4,000-6,000 a m p s ) , then the oxygen porous anode process could present t h e most economical method f o r producing OF2.

52

V I 1 Bibliography

A. Fluorine-Caustic References

A-1. Arkell A e t a l , "Basic Chemical Research fo r Advanced Storable Liquid Oxidizers", 1966, AD 373,568.

A-2. Brown R. A. e t al, "Production of Oxygen Difluoride" U.S. Pat. 3,367,744, 1968.

A-3. Borning A. H. and K. E. Pullen, Inorganic Chemistry, E, No. 8, 1791, 1969.

A-4.

A-5.

A-6. Engelbrecht A. and H. Atmanger, J. Inorgn. Nucl. Chem.

Cady G. H . , J. Amer. Chem. SOC. - 56, 1647, 1934.

Cady G. H . , J. h e r . Chem. SOC. 5'7, 246, 1935.

- 2, 348, 1956.

A-7. Holzmann R. T., J. Dvorak, T. Hirata "Synthesis and Characterist ics of High Performance Sol id Rocket Oxi- d i zers" , 1961.

A-8. Koblitz W. and J.J. Schumacher, Z phys chem. B25, 283, - 1934.

A-9. Lebeau, P. and A. Damiens, Comptes Rendus - 188, 1253, 1929.

A-10. Rocketdyne "Research i n High Energy Oxidizers", 1965.

A-11. Rohrback G. H. and G. H. Cady, J. Amer. Chem. SOC. 69, 677, - 1947.

A-12. Rohrback G. H. and G. H. Cady, J. Amer. Chem. SOC. 70, 2603, - 1948.

A-13. Ruff, 0. and W. Kwasnik, Angew Chem. - 48, 238, 1935.

A-14. Ruff, 0. and W. Menzel, Z anorg a l l e g chem. 190, 257, 1930.

A-15. Ruff, 0. and W. Menzel, Z anorg a l l e g chem. 198, 39, 1931.

A-16. Schnizlein, 3. G. e t a1 3. Phys. Chem. - 56, 233, 1952.

A-17. Sicre , J. E. and H. J. Schumacher, Angew Chem. 9, 266, 1957.

A-18. Swinehart, C. F., "Final Summary Report fo r Rocket Pro- pe l l an t Oxidizers" Harshaw Chem. Co. 1950.

A-19. Wartenberg, von H. and G. Klinkott, Z anorg a l l e g chem. 193, 409, 1930.

53

B. Fluorine Oxygen References

b

I

B-1. Aoyama S. and S. Sukuraba, J. Chem. SOC., Japan, 2, 1321, 1938.

B-2. Aoyama S. and S. Sukuraba, J. Chem SOC., Japan, 62, 208, 1941.

B-3. Goetschel C.T., "Fundamental Research on Advanced Oxidizers" Quarterly Report #15, Aug. 1966, AD 803, 204L.

B-4. Grosse, A. V. , A.G. Streng, and A.D. Kirshenbaum, J. h e r . Chem. SOC., 83, 1004, 1961.

B-5. Lawless , E.W., and I . C . Smith, Inorganic High Energy Oxi- dizers . Marcel Dekker Inc., N.Y. , 1968.

B-6.

€3-7.

Kirshenbaum, A.D., Zeit F. Chem. 9, No. 4 , 121, 1969.

Kirshenbaum A.D., A.V. Grosse, J.Amer. Chem SOC., 81, 1277, 1959

B-8. Kirshenbaum A.D., e t al Grosse, J. h e r . Chem. SOC. 8 l , 6398, 1959.

B-9. Pennsalt, "Synthesis of Inorganic Oxidizers", AI? 04(611)- 851e, Quarterly Report No. 5, 1964.

B-10. Ruff 0. and W. Menzel, Z anorg a l l e g chem. 211, 204, 1933.

E-11. Ruff 0. and W. Menzel, Z anorg a l l e g chem. - 217, 85, 1934.

B-12. Streng A. G., and A. V. Gross, "Addition and Subst i tut ion Products of Oxygen Fluorides" 2nd AnnQal Report, 1962, AD 273, 732.

B-13. Streng A.G. and A.V. Gross, "Addition and Subst i tut ion Products of Omgen Fluorides", 4th Annual Report 1964, AD 435, 838.

54

C. Direct Elec t ro lys i s References

C-1. Allen D. R. Chem. Ehg. - 67, 216, Sept. 19, 1960.

C-2. American O i l Co., "The FO Radical and High Pressure Re- act ions of N2F2," F i r s t Annual Report , 1964, AD 432 057.

American O i l Co., "The FO Radical and High Pressure Re- act ions of N2F2," Second Annual Report 1965, AD 459 596.

American O i l Co., "The FO Radical and High Pressure Re- act ions of N2F2," Third Annual Report 1966, AD 482 876.

Childs W.V., "Electrochemical Production of Oxygen D i - f luoride", U.S. Pat. 3,461,049, 1969.

C-3.

C-4.

C-5.

C-6. Donohue J . A . et al, Advances i n Chem Ser. No. 54 1965 p. 192.

C-7. Donohue J . A . e t al, "Elec t ro ly t ic Production of Ozone," U.S. Pat. 3,256,164, 1966.

C-8. Donohue J .A. and W.A. Wilson, "Electrolyt ic Production of Oxygen Difluoride", U.S. Pat. 3276,981, 1966.

C-9. Engelbrecht A e t a l , Monatsh Chem. 95, 633, 1964.

C-10. Engelbrecht A and E. Nachbauer, Monatsh Chem. 90, 367, 1959.

(2-11. Grubb W.T., "Electrolyt ic Method of Producing Fluorine and Fluorine Oxide", U.S. Pat. 2,716,632, 1955.

C-12. K i s a k i H et al, Denki Kajaku, 34, 24, 1966.

C-13. Lebeau P and A Damiens, Compt Rend, 185, 652, 1927.

C-14. Nagase, S., H. Baba, R. Kojima, Bull Chem Soc. Japan, 35,

-

1907, 1962.

C-15. Nagase, S., H. Baba, R. Kojima, J of Chem SOC. Japan, Ind. Chem, Sect. - 64, 2124, 1961.

c-16. Rogers H.H. e t al, J. Electrochemical SOC. 111, 704, 1964.

C-17. Rogers H.H. and J . H . Johnson, "Research i n Fluorine Chemistry Quarterly Progress Report f o r Period Ending Sept. 15, 1961", Rocketdyne.

C-18. Schmidt, H. and H.D. Schmidt, Z anorg alleg. Chem., 279, 289, 1955.

C-19. Simons J . H . e t al, J. Electrochemical Society, 95, 59, 1949.

55

L

D. General References

D-1 . D-2.

D-3.

D-4.

D-5

D-6.

D-7.

D-8.

D-9

Chemical Propulsion Information Agency, "Oxygen Difluoride" , Liquid Propulsion Manual, U n i t 28.

Donohue, J . A . , and F.S. Jones, Anal. Chem., - 38, 1858, 1966.

Fox W.B., and R.B. Jackson, "Ovgen Difluoride", Kirk and Othmer, Encyclopedia of Chemical Technology, 2nd edi t ion V O ~ . 9, 631-635, John Wiley and Son Inc.

George, H. , "Properties of Selected Rocket Propellants",

N.Y., 1966.

1963, AD, 444, 642.

Leech H.R., Mellor's Comprehensive Treatise on Inorganic and Theoretical Chem. Supp. I1 Part I. Longmans Green & Co., Inc. N.Y. p. 186-93, 1956.

Oxygen Difluoride" , Product Data Sheet General Chemical I 1

Division, All ied Chemical Corp., Morristown, N. J.

Ryss I . G . and F. Haimson, The Chemistry of F2 and i ts In- organic Compounds 162-164, t ransl . by AEC, Feb. 1960.

Sekits D.F. "Design Handbook for Ovgen Difluoride", Nov. 1963, AD 350,631.

Streng A.G., Chem. Rev. 63, 607-624, 1963.

D-10. Yost D.M. and G.H. Cady, Inorg. Syn. V o l . 1, 109, 1939.

CASE FILE - NASA Report CASE FILE lrPYIh COW! .d November 1970 ... those for the fluorine-caustic...

Documents

Transcript of CASE FILE - NASA Report CASE FILE lrPYIh COW! .d November 1970 ... those for the fluorine-caustic...