Cu2O thermite compatibility studies by x-ray photoelectron and x-ray induced auger spectroscopy

Journal of Hazardous Materials, 5 (1982) 297-308 Elsevier Scientific Publishing Company, Amsterdam - Printed in The Netherlands

Al/Cu20

THERMITE

COMPATIBILITY

STUDIES BY X-RAY PHOTOELECTRON

297

AND X-RAY INDUCED

AUGER SPECTROSCOPY

P.S. WANG, L.D. HAWS and W.E. MODDEMAN Mound Facility*,

Monsanto

Research

Corporation,

Miamisburg,

Ohio 45342

A. RENGAN Research

Institute,

University

of Dayton,

Dayton,

Ohio 45469

ABSTRACT

The surface

chemistry

of A1/Cu20

thermite

in powdered

pellets was studied before and after accelerated spectroscopy energies

(XPS) and with x-ray

mixtures

and in pressed

agings with x-ray photoelectron

induced Auger spectroscopy

of the XPS or XAES signals were observed

The kinetic

(XAES).

for Al Is, Cu 2p

0.l 2P3,2'

l/2'

cu L-MM, 0 Is, Al K-LL, C Is and Cu 3s. The A1203

film thicknesses on Al metal surfaces were deduced from signal inten+3 and Al in XPS of Al 2s and from Al K-LL XAES signals. The

sity ratios of Al oxide thicknesses

on Al powders

to form the thermitic

with Cu20 at room temperature, tion of the Al and increased powders

from the manufacturer

layer was found to protect for several

months

Stearic

the Al fuel; further aging of these pellets

acid, the organic

by unpaired

impurities

of thermite

was found to double the oxide layer.

showed a negligible

The CuO impurity

broadened

induced oxida-

Hot pressing

change in composition

This oxide at 180°C

and thickness

of the

The bulk of the pellet was found to be stable and of good quality

surface oxide.

signal.

composition,

the oxide film thickness.

at 425°C to form pellets

Mixing Al

were measured.

additive

electron

were removed

to Al powders,

in Cu20 was observed spin-spin

was detected

in the Cu 2p

interaction

during the high temperature

by the C 1s

signal which was

3/2 in the 0.1~~ dg orbital; pressing

both

operation.

INTRODUCTION The A1/Cu20

thermite

heat is released

2Al + 3Cu20

is a valuable

chemical

for each gram of A1/Cu20

heat source because

mixture

according

580 cal of

to the reaction:

A1203 + 6Cu.

(1)

*Mound Facility is operated by Monsanto Research Corporation Department of Energy under Contract No. DE-AC04-76-DP00053. 0304-3894/82/0000-0000/$02.75

0 1982

for the U.S.

Elsevier Scientific Publishing Company

298

Aluminum

powder,

being very sensitive

to oxygen or other oxidizers,

grows a film of A1203 on its surface may or may not protect under various in addition

the aluminum

fabrication

and storage

to the problem

problems

A1203 has much higher melting

quality

To increase pressed

of various

fabrication,

and storage

surface

is indeed critical

metal.

process

The main interest

inevitably

We are concerned

has on the quality

powders

to

by x-ray photoelectron

particles

in our thermite

promotes

Stearic

acid has been pro-

may also be informative. of aluminum

fabrication,

(XPS), or electron

it is

a certain

and surface

the thicknesses

(ESCA), and x-ray induced Auger electron

need to be

Therefore,

on the size and geo-

as a lubricant

during production,

spectroscopy

powders

with the degree of in-

of the thermite.

by the manufacturer

of this study is to measure

oxide layers on aluminum

thermite

to 425°C or above, depending

the role played by this "impurity"

analysis

to 660°C for aluminum

shapes for special use.

This heating

in aluminum

fluence this heating added to aluminum

the

of the fact that

of the oxide film on the aluminum

desired

to heat the powders

degree of oxidation

because

and compatibility.

metry to be consolidated.

tector;

(205OOC compared

the energy output per unit volume,

to pellets

necessary

can be anticipated

point

during the stages of production, the material

fuel)

If the layer grows too thick,

conditions.

the thickness

Consequently,

metal).

This oxide film

to the air.

(which is our thermite

inside

of losing too much of the fuel and thus lowering

compatibility

heat output,

after exposure metal

immediately

spectroscopy

spectroscopy

and storage for chemical

(XAES).

(ref. 1,2)

EXPERIMENTAL Sample preparation

and agin%

Theoretically

stoichiometric

were produced

by dry mixing

oxide in a V-blender and Cu20 powders cles.

Thermite

of theoretical

Reynolds

pellets

Aluminum

a 400-mesh

about 6 mm in diameter,

density were pressed

(11 wt % Al and 89 wt % Cu20)

XD28 aluminum

for 1 hr (ref. 3).

were sieved through

under dry nitrogen Thermite

thermite mixtures

from powders

flakes and Cerac "pure" cuprous flakes were used as received

screen to remove oversized about 2 mm in height, with preheated

at 425°C and 12,000 psi for 15 min

mixtures,

thermite

glass tubes separately. were aged at desired

pellets,

and aluminum

TWO environments

temperatures

and at 90%

graphite

dies

(ref. 3,4). powders

were used:

for desired

parti-

were sealed into

dry air and argon.

Samples

periods.

Spectrometer The XPS spectrometer with the accompanying detail

elsewhere

is a modified

AEI ES-100

argon ion sputtering

(ref. 5).

pumps, the vacuum achievable

The system

instrument.

capability

This spectrometer

has been discussed

in more

is pumped by two 250-liter/see turbomolecular -8 in the sample chamber being 10 The chamber torr.

299 reaches baking

temperatures

nitrogen

cryostations.

of more than 15O'C.

by using a titanium

in the system by baking,

Reactive

sublimation

vapors

can be minimized

pump, and by two liquid

The anode used for all the XPS measurements

was silicon.

RESULTS AND DISCUSSION Calculation

of surface oxide and carbon

The oxide thickness the following

d

0

(do) on aluminum

equation

film thicknesses particle

surfaces

can be calculated

from

(ref. 6-8):

= x0 In I(Io/Im) (I;/Iz) + 11

(2)

where

A is the mean free path of an Al 0 electron. are measured spec0 2 3 IO’ Irn troscopic signal intensities for Al 0 and Al metal, respectively, for our samples 2 3 m cc I and I are intensities for "infinitely" thick samples of such materials. The 0 m carbon film thickness (dc) can be calculated from (ref. 7,8):

d

I hc In {l - +

=c

}

(3)

C

m is the mean free path of an electron from carbon, and I and I are the c C C carbon signal intensities of the thermite sample and of the sample of "infinite"

where

h

carbon thickness,

respectively.

The mean free path of a photoelectron the material attenuation

through which

it passes.

of the photoelectron

A least-squares

fit model which has been developed

KLL Auger with

from the curve 1385 eV kinetic

on its kinetic

energy and

et al. (ref. 9) have studied

from Al 20 3 (kinetic energy

over 1600 eV to cover our kinetic was extrapolated

is dependent Battye

the

from 157 to 1404 eV).

was used to extend the data

energy range and our mean free path of A1203

(ref. 10).

A x0 of 16.5 i was obtained

for Al

energy and 18.1 i for Al 2s ESCA at 3615 eV kinetic

energy. To evaluate tween X

c

hc through

a carbon

of a photoelectron

layer, Wagner

to its kinetic

has shown that the relation

energy E is:

xc = KEp

(4)

where K and p are constants lengths

(ref. 11).

in stearic acid multilayers

respectively

(ref. 12).

Henke has shown that the attenuation

are 60 i and 90 i for 750 eV and 1350 eV,

This gives K = 0.62365

and p = 0.68983

Ac = 0.62365E0.68g83

xc

be-

and

(5) 0

was found to be 94.7 A for C 1s ESCA electrons.

300

To measure

1: (the total intensity

molecular-weight and stearic

carbon-containing

compounds

thick carbon layer) high-

such as candle wax, Apiezon

acid were layered on a 5-mm x 5-mm aluminum

Proof of "infinite" KLL Auger aluminum counts-eV/sec

thickness

the signals

observing

of an aluminum

completely

An averaged

the signal.

the

value of 890

oxide layer of infinite

thickness

(10)

foils were aged at 400°C for 12 hr to con-

to aluminum

Values

of any metal signals.

by scanning

from these measurements.

layers of 5-mm x 5-mm aluminum

vert the surfaces

grease,

foil for measurements.

of the carbon layer was obtained

lines without

was obtained

To measure several

of an infinitely

oxide, which was proved by the absence

of 5725 and 573 counts-eV/sec

were obtained

for Al

KLL Auger and Al 2s ESCA, respectively. The total intensity by measuring

thick Al metal sample

of an "infinitely"

a fresh Al surface

(I:) was obtained An oxide build-up

cleaned by argon ion-etching.

curve was established

in Al 1s ESCA to calibrate possible surface oxide growth m The values of I are 4298 and 501 counts-eV/sec for during the measurements. m KLL Auger and Al 2s ESCA, respectively.

Surface

analysis

It is logical background

ESCA spectra

of aluminum

Carbon

powders

of (a) aluminum

alone as

Fig. 1 gives

flakes aged at 180°C and (b) thermite by Si K, under the same

conditions.

1s signal

- stearic

could be deconvoluted

acid in aluminum

into a doublet

in more than one chemical

carbon of a straight-chain 2a).

structures

for further study of thermite mixtures.

aged at 18O'C; both samples were irradiated

spectrometric

present

powders

to study the surface

knowledge

the overall powders

of aluminum

form.

hydrocarbon

The ratio of the straight-chain

carbon at 286.6 eV was 14:l.

powders.

structure,

This carbon photopeak

which is characteristic

of carbon

The two peaks noted can be attributed and carbon of a carboxylate

hydrocarbon

By comparing

the signal of the specially

prepared

to be from the stearic acid {CH3(CH2)16C02H}

to the aluminum

powders by the manufacturer.

stearic

(carbon 1s photopeak)

was subsequently

identified

(Fig.

at 284.8 eV to the carboxylate

this signal

standard

species

to

acid sample

The theoretical

with

(Fig. 2b), it added

ratio of the two

peaks should be 17:l. 'Ihe amount of stearic ture, as expected, temperature.

acid was found to be less at the higher

because

of removal by higher vapor pressure

Fig. 2c shows that only 75% of the stearic

sample after aging at 180°C in dry air for 40 days. at the higher

storage

temperature

the amount of carbon measured

tempera-

at the higher

acid remains

in the

The amount of carbon measured

in argon was found, in general,

at the same temperature

storage

for storage

to be less than in dry air.

Al IS

Flakes

Kinetic Energy, eV ___t

from aluminum

A1+3 KLL

AI-KLL

Al+3 KLL .

Fig. 1. Overall XPS spectra of (a) 180°C aged flaked aluminum and (b) 180°C aged thermite powder specimens. Both spectra were excited by SiKol radiation.

(b) Thermite Powder

(a) Aluminum

metal

302

I

264.8

eV

C-c, C-H of R Group

1 eV -

-

Binding Energy, eV

Fig. 2. ESCA carbon Is photopeak of stearic acid (a) aluminum flakes stored at room temperature, in: (b) stearic acid sample, and (c) aluminum powder stored at 180°C for 40 days in dry air.

303 Surface

change

powder before

in

aluminum

powders

aging is approximately

and tended to be destroyed

by aging.

alone.

carbon

Is signal intensity

aluminum

this period

during

Residual

for 40 days, powder

aging.

of added stearic

Fig. 3 shows such a change in the Table 1 gives results of layer thick-

stored at room temperature,

for the same samples.

The thicknesses

layers are also shown in Table 1.

Room Temperature

A6

Fig. 4 shows

powder stored at 180°C

stored at 180°C for 91 days, and flamed foil. peaks

acid

carbon after aging could be attrib-

from Al KLL Auger and Al 2s ESCA data.

KLL Auger of powder

the 2s photoelectron

layer on the aluminum

50-60 i thick because

uted to the x-ray tube or the atmosphere.

nesses during

The carbon

days

Kinetic Energy, eV _) Fig. 3. Carbon 1s XPS of aluminum powders aged at 180°C in air for various periods.

Fig. 5 shows of these oxide

304 TABLE 1 Aluminum-KLL environments

Auger and 2s ESCA results

for aluminum

powders

aged in different

do (A)

Sample d_ (%

KLL Auger

Aluminum powder in air at room temperature

64.4

8.7

11.1

Aluminum powder in air 40 days at 180°C

56.8

11.6

12.4

Aluminum powder in air 91 days at 180°C

62.3

13.0

13.0

Aluminum powder in air 136 days at 180°C

41.8

13.0

14.0

Aluminum powder in argon 40 days at 180°C

10.6

9.1

11.2

Aluminum powder in argon 91 days at 18O'C

16.9

9.1

10.8

Aluminum powder in argon 136 days at 18OOC

26.7

12.0

12.0

Aluminum

powders

from the manufacturer

Al 2s

have an oxide layer of about 9 i as

expected because of exposure to the air during storage. This layer increased to 0 0 about 12 A when aged about 40 days in air at 180°C and to 13 A in 3 mo. After 3 mo there seems to be no further of the sample tube

no oxygen was present, increase

storage

in the oxide layer

experimental

error.

because

The 2s signals

air).

(about 9 i), but the result derived

for 136 days could be an

The deviation

there is a certain

On the lower-energy

side of both the KLL and 2s signals, of the surface plasma.

Thermite

mixture

The overall Comparing

surface

higher

than

of results

degree of error

technique.

weak signal because to the main signals.

since

to two reasons:

are not well resolved;

in curve deconvolution

of the small size As expected,

from Al 2s ESCA are slightly

but show the same trend.

from KLL to 2s can be attributed

(2)

possibly

under argon at 180°C for 40 and 91 days showed no

The results

those from the Auger process

(1)

increase,

(which could not supply additional

The correction

there exists a

This needs to be corrected

term is smaller

for 2s than for KLL.

analysis

XPS spectrum

of thermite

Fig. la to lb, it is obvious

powders

aged at 180°C is shown in Fig. lb.

that in thermite

powders, Cu 2p l/2' cu 2P3,2' and Cu L M of Cu20 were detected. Unfortunately Cu 3s and Cu 3p happen 3 4,5 M4,5 to be at the same position as Al 2s and Al 2p, respectively.

305

1383.5

ev

1390.8

eV

Kinetic Energy, eV w Fig. 4. Al-KLL spectra of: (a) aluminum flakes stored at room temperature, (b) aluminum flakes after mixing with CuaO to form thermites at room temperature, (c) thermite powders aged at 180°C for 91 days in dry air.

306

1383.5

eV

Al+3

b)

Kinetic Energy, eV

w

Fig. 5. Al-KLL Auger spectra of: (a) Al/Cu20 thermite pellet after pressing at 425"C, (b) pellet aged at 18O'C for 27 days in dry air.

307

The carbon Is signal from thermite carbon Is signal

from aluminum flakes.

powder

flakes.

After the sample

signal

from aluminum

heated

to 18O"C, the signal was reduced

a small degree of oxidation removing

stearic

occurred

during

figure,

the mixing

This suggested

and the heating,

that

thereby

flakes before mixing with Cu20,

powders

to form thermite

aged at 180°C in dry air for 91 days.

that oxide

film thickness

increases

(b) powder

Just from this

after mixing and further

during aging.

Results Table 2.

to that

acid.

it is obvious

increases

of the

(either in dry air or argon) was

flakes after mixing with Cu20 at room temperature

and (c) thermite

one-fourth

almost to background.

Fig. 4 shows Al-KLL Auger of (a) aluminum aluminum

is approximately

Also, it is skewed when compared

derived

from the Al-KLL Auger for thermite

The oxide film of the powders

powders

at room temperature

are tabulated

in

was found to be about

TABLE 2 Aluminum-KLL conditions

Auger results

for aluminum/cuprous

oxide thermite

do (i,

Sample Powder in air at room temperature

10.8

Powder in air 40 days at 18O'C

21.4

Powder in air 91 days at 180°C

24.8

Powder in air 136 days at 180°C

23.8

Powder in argon 136 days at 18OOC

22.1

Air, room temperature,

22.3 pellet

Air, 27 days at GO'C,pellet surface

25.4

Air, 27 days at 6O"C, fractured pellet

23.1

Air, 27 days at 18O"C, pellet surface

25.7

Argon, 131 days at 18O"C, pellet

28.1 surface

Argon, 131 days at 180°C, fractured

23.4 pellet

aged at various

11 A, which means that just mixing creates

an oxidation

layer increases

reaction

aluminum

powder

that increases

and Cu20 at room temperature

the oxide layer by 2 i.

The oxide

to about 21 i for the first 40 days at 180°C in air and levels

off after 3 mo at 24-25 A.

In argon at 180°C for 136 days, the oxide layer was

found to be about 22 i. Fig. 5 shows the Al-KLL Auger of a thermite

pellet

(b) after aging at 180°C for 27 days in dry air. that aging of the pellet ness; but comparing induces

shows an insignificant

Fig. 5a to Fig. 4b proves

some oxidation

and increases

In order to press thermite up to 425OC.

This increased

aging of the pellets

(a) after pressing,

Comparing

change in the oxide layer thickthat pressing

of the pellet

it was necessary

to bring the temperature

the oxide film from about 11 to about 22 i.

caused little increase As expected,

the pellet was found to be thinner

indeed

the oxide layer considerably.

pellets,

than 6 i in 131 days at 18O'C.

and

(a) and (b) indicates

in oxide layer thickness

Further

- no more

the oxide layer in the bulk part of

than that on the surface.

The results

are

shown in Table 2.

REFERENCES 1 K. Siegbahn, C. Nordling, A. Fahlman, R. Nordberg, K. Hamrin, J. Hedman, G. Johansson, T. Bergmark, S. Karlsson, I. Lindgren and B. J. Lindberg, Atomic Molecular and Solid State Structure Studies by Means of Electron Spectroscopy, Almquist and Wiksells, Uppsala, 1967. 2 K. Siegbahn, C. Nordling, G. Johansson, J. Hedman, P. F. Heden, K. Hamrin, U. Gelius, T. Bergmark, L. Werne, R. Maune and Y. Baer, ESCA Applied to Free Molecules, North Holland-American Elsevier, Amsterdam, New York, 1969. 3 L. 0. Haws, M. D. Kelly, J. H. Mohler and A. Latkin, Consolidated Al/&20 Thermites, presented at the 6th Int. Pyrotechnic Seminar, Estes Park, Colorado, July 17-21, 1978, MLM-2531(OP). 4 A. Gibson and J. R. Marshall, Monsanto Research Corporation, private communication. 5 T. M. Wittberg, J. R. Hoenigman, W. E. Moddeman, C. R. Cothern and M. R. Gulett, J. Vat. Sci. Technol., 15(1978) 348. 6 T. A. Carlson and G. E. McGuire, J. Electron Spectrosc. Relat. Phenom., 1(1972/1973) 161. 7 P. S. Wang, L. D. Haws, W. E. Moddeman and A. Rengan, Aluminum Particle Surface Studies in Al/Cu20 by Electron Spectroscopy for Chemical Analysis and Auger Electron Spectroscopy, MLM-2664 (November 9, 1979), 23 pp. 8 W. E. Moddeman, A. Rengan, P. S. Wang and L. D. Haws, Proceedings of the 10th Symposium on Explosives and Pyrotechnics, San Francisco, California, February 14-16, 1979. 9 F. L. Battye, T. Liesegang, C. G. Leckey and J. G. Jerkin, Phys. Lett., 49A(1974) 155. 10 A. Rengan and W. E. Moddeman, X-ray Photoelectron Studies on the Thermite Mixture of Aluminum and Cuprous Oxide, UDRI-TR-77-30, University of Dayton Research Institute (19771 _ 11 C. D. Wagner, Anal. Chem., 49(19771 1292. 12 B. L. Henke, J. Phys. (Paris), C4(1971) 155.