THE INFLUENCE OF THE GAS ATMOSPHERE AND COMPACTING PRESSURE ON REACTIONS IN THE SOLID STATE

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP78-04861A000400030012-3
• Release Decision: RIPPUB
• Original Classification: K
• Document Page Count: 31
• Document Creation Date: December 20, 2016
• Document Release Date: June 6, 2006
• Sequence Number: 12
• Publication Date: July 1, 1955
• Content Type: REPORT

Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
THE INFLUENCE OF THE GAS ATMOSPHERE AND
COMPACTING PRESSURE ON REACTIONS
iN THE ,SOLID STATE.
Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
 Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
25X1
TBE INFLUENCE OF THE GAS ATMOSPHEPE AND COMPACTING
PRESSURE ON R+;ACTIONS IN THE; SOLID STATE,
G. Henrich.
Zeitschrift fur Elektrochemie, .~. 3 (1954) 183-196.
(From German)
To examine the influence of foreign gases and possible contact
effects, on the ruction velocities of solids, an apparatus is
described in which the progress of`reactions with time in the solid
state can be followed in specific gas atmospheres. This apparatus
has been used to investigate the temperature dependence of a number
of spinel reactions in nitrogen as well as the influence-of the
compacting pressure, the particle size, the degree of mixing of
different foreign gases (O,, H2,-00a, SOaand NH.) and vacuum on
the reaction velocity.
The compacting pressure acts in three, clearly distinct
different ways, which to some extent overlap. The particle size
( of A120):in the system ZnO/Al0has practically no influence
on the reaction velocity .0 promotes the reaction MgO + Cr 0
very considerably, even in slight traces, which is explained by
the intermediate formation, of MgCrO . Whilst ? 0a has. no effect in
the System MgO/Al2'0 'the reaction;`MgO + Fe0a is. markedly
inhibited by 0 , for which the repression of the thermal.
dissociation of Fe200 may be made responsible. H and NHs
exercise no influence in the system Mgo/Al 03. Whereas CO2 is
without visible.effect in the system MgO/AlO , this gas atmosphere
in the system ZnO/Al Osappreciably accelerates the reaction
velocity. SO considerably promotes the reaction in the system
MgO/AlO No definite proof was obtained of any variation of
the reaction velocity in a vacuum,
Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
 Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
In a large number'of investigations, Hedvall (15-22), Forestier
8-14, 30, 31-, 37), Huttig (24, 25), Fricke (4) and other authors
3, 38) have dealt with the problem of the influence exerted on
reactions in the solid state by the compacting pressure and foreign
gases. A reliable interpretation of the experimental results is
obviously prevented by the difficulty of varying only one of the
different possible experimental conditions at a time. As these
relationships have not yet been adequately clarified, the present
investigation is intended to provide a contribution towards this end,
The systems which have been selected for this are distinguished for
their unmistakable chemical reactions and for the simple methods of
determination.
Those requirements are sufficiently closely satisfied by the
spinel systems consisting of the components MgO, ZnO, and CdO, as well
as Cr 0 , Al 0 and Fe 0 . All the systems obtained with these
2 3 2,3 2 3
components have been investigated, with the exception of zinc ferrite
and cadmium ferrite.
A. Start substances and experimental method.
I.SStartin substances.
. _ Ma nesiLUn oxide.
The basic nhaghesium carbonate p.a. of Merck was ignited tightly
pressed in a crucible furnace at first for 1 hour,at,.500- 600?C and
then for 24 hours at 1100-1200?C. . Examination for impurities showed
the absence of detectable quantities of halogen, sulphate and heavy
metals and less than 0.l7o of sodium. Under the microscope with a
magnification of 500 smal.l, transparent, often regular crystals
with a diameter of about 24i were to be, seen.
2. Zinc oxide.
Zinc oxide p.a. of Merck, when examined for impurities, showed
substantially the same results as magnesium oxide. For sintering,
the product was pressed tightly in a clay crucible and ignited for
8-10 hours at 1100?C, during which it sinter(-,d,-its volume being
reduced by about 1/4. Microscopic determination of particle size gave
a mean diameter of 26i .
CadmiL+rn oxide.
Cadmium acetate p.a. of Merck was precipitated with ammonium
carbonate; the CdCO was filtered, dried and slowly decomposed in a
crucible furnace. a The resulting CdO was ignited for 8 hours at
800?C, sintering together to form a chocolate brown core. A higher
sintering temperature was not selected because of the appreciable
evaporation of CdO which occurs at 900?C according to equation (14)-
For References, see end,
Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
 Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
?+v Alumiriurn oxide.
Merck',s Broclouan aluminium oxide was used as starting material.
In order to remove some alkali content, which it possesses according
to manufacturer' ss statements, it was boiled for 1 hour with 2n
hydrochloric acid with vigorous agitation,, filtered at the pump,
washed and dried.. It was then sintered for 12-15 hours at 1100?C.
Undor the microscope, all particle sizes between 5 and 100?
were found in fairly uniform distribution.
"Pe 0 according to Brandt" was used as starting substance,
and this~,vas united in a clay crucible for 10 hours at 1100?C.
The colour changed from reddish brown to bluish gray; the product
sintercd to about 1/3 of its original volume to a solid co'r`e.
After grinding in a mortar, black particles with a mean diameter
of 4-7? were mostly to be seen when examined under the microscope.
6. Chromium oxide.
Purest" chromium oxide of Merck was ignited. for 10 hours at
11500C in a current of purified nitrogen. Externally, no
sinterin was observed. The colour became somewhat darker. The
mean particle size, determined under the microscope was about 2p.t.
IIPreparation of the mixtures and compressed
com acts.
Mixing of the substances, wei0 ed out in the desired
proportions, took place in a tt lltimix" made by the firm of Braun.
This apparatus was found to be extremely useful in the present
work, and can be highly recommended wherever there is a need for
mixing powders as intimately and homogeneously as possible,
.A certain, truly not very great disadvantage of the "Multimix", is
its tendency to grind, which for some experiments cannot be
disregarded,
2. ,. Pressing.
Pressing of the compacts was effected in a hydraulic
laboratory press..- "Thd cylindrical moulds used had a diameter of
11.0 mm,, by means of which compacts up to pressures of 12.5 t/cm
could be made. On each pressing stroke, there was a waiting time
of one minute to allow equilibrium to become established.
Since it was found in the first pressing experiments that the
compacts usually broke in two when removed ,from the press, all the
compacts ;-Jere broken through the c(: ntre and were subjected to
reaction as half compacts.
The ,vi) i ht of a compact was about 7-8. Mval.
The compacting pressure is indic_,ted in the individual reactions.
Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
 Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
III, -Apparatus'.
All the reactions were carried out in a Silit rod furnace
(current 10 amp., power 2 kW). t,. Silit rods A (see Figure 1)
were used as heating rods. The regulating thermocouple C which
was in a porcelain tube D, was introduced into the furnace from
the end, so that the free junction E was directly adjacent a
Silit rod, the regulation of the temperature being Owls free
from inertia as possible. The furnace temperature could be kept
constant to t 5?C. The hard Porcelain heating tube B (internal
diameter 30 mm, length 650 mm) was glazed at both ends to glass
tubes F and G. The glazed-on tubes terminated at either end in
a ground sleeve S I and S 2. The ground inner piece connected to
S I carried a glass tube H (10 nun diameter) projecting into the
ground joint and having glazed to it a porcelain tube I, which
served for receiving the measuring thermocouple. The junction of
this thermocouple lay in the middle of the heating tube somewhat
above the centre. Between S I and the glazed joint was a
shortened pressure gauge J and the gas inlet K provided with a
cock. Figure 2 shows the ground - joint attachment for S 2.
The porcelain tube L could be pushed in the guide support M, a
piece of vacuum tubing N serving as gas seal. This porcelain
tube L served to carry a 3 mm thick iron wire 0, to the front end
of which was rivetted a round scraper P, 12 mm in diameter.
The gas seal between L and 0 was also provided by a piece of
vacuum tubing. The glass wall U was fused as flat as possible
to provide a good view into the interior of the tube.
IV~Gas_~urif'ication~
Most of the experiments were carried out in an atmosphere of
nitrogen. In some cases, ordinary commercial nitrogen containing
3-0 oxygen was used, in others a nitrogen purified from oxygen
and having an oxygen content of less than 10-2%.
Purification of the nitrogen was carried out by the method
of Meyer and Ronge (32). For adsorbing any oil vapour which
could interfere with the catalytic effect of the active copper., a
vessel containing silica-gel was inserted between the boni'b and the
Meyer-Ronge tower (referred to in what follows as the M-R tower).
After the oxygen absorption, the gas passed through a
washbottle containing concentrated sulphuric acid, which served at
the same time as bubble counter; it then passed through a
Fresenius drying tower containing silica-gel, and finally thrqu`ch
a drying tower filled with alternate layers of phosphorus
pentoxide and glass wool.
In reactions in a current of hydrogen, the gas was also
purified in the manner described, whilst all the other gases entered
the sulphuric acid bottle directly, but otherwise massed through the
same drying process; only in the case of ammonia, _ drying was done
by means of soda lime.
V. Carryin out the exerimc nts.
The experiments were all carried out as follows: The compacts,
broken in two in the middle, were placed with the fracture on the
Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
 Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
support (a 20 x 70 mm piece of asbestos card 6 mm thick) and then
the su)-port with the compacts was pushed just behind the glazed
joint at R: The temperature at that point was 150-200?C, the
substance being thereby freed from any superficial moisture.
After closing the furnace, the around joint attachment T was
placed in position and the gas was passed through for about - to 1
hour, whereupon by~means of the scraper, the support with the
compacts was introduced into the hot zone of the furnace.
At the appropriate times, whilst observing through the glass
window U, a compact could be drawn off the support and placed in the
ground joint attachment T. After completion of the experiments,
the ground joint attachment was removed and the compacts analysed
by the method described in the next section.
It was in this way possible to allow the entire reaction to
proceed to conclusion in a definite gas atmosphere.
VI. Aral tical rn tho.
In the analytical investigation of the compacts after the
reaction for determining the percentage conversion it was important
to determine, in addition to the acid-insoluble Me20a and spinel,
also the acid-soluble MoO . As preliminary experiments had shown
that even highly ignited MeO, if powdered finely enough, dissolves
quantitatively in 0.1 n -hydrochloric acid when heated on the water
bath for half an hour, the following method wLs adopted foxr analysing
the compacts:
A suitable weighing, containing about ;3 Mval free IMO, was
heated in a 300 ml Erlenmeyer flask for 30 minutes with 50 ml 0.1
n-hydrochloric acid on a well boiling water bath, with frequent
shaking. It was then filtered at once hot from the insoluble
residue (spinel + Me 0 ) and the excess hydrochloric acid in the
filtrate was titratel ?ack with 0.1 n-sodium hydroxide.
To prove the accuracy of this method, 9 reaction determinations
by this method were carried out on an intidte mixture of MgO and
Al 0 , which had. been heated for 13 hours at 1100?C. A mean
value of 60.2dd was obtained with a maximum error of t 0.5%
This relatively large error must be accepted in this reaction
determination. As shown by a few preliminary experiments, it is
partly due to the fact that a certain uncontrollable quantity of
acid is consumed by the spinel and by Mea03. There is, furthermore,
the possibility that some acid may be lost in digesting and filtering.
The content of MeO in the starting mixture, the so called "zero
value". was therefore not calculated from the weighed quantity, but
was.titrated in the same way as the compacts after reaction.
An error in the reaction determination of t 0.5''o is furthermore
substantially less than the reproducibility of the reaction, which is
Me0 denotes the divalent oxide and Me20a the trivalent oxide.
Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
 Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
generallj not .more than t 1 ~ and is often -.torso than 2%,
The fluctuations are particularly high at the commencement of the
reaction, where the scatter of the values is not infrequently
about t 5,. This is easily understood since at the conw.encement
of a reaction, the spinel is in a very active state, so that it
may be dissolved by the acid to some extent.
B. Ex}?er~nental results.
I. Reaction MgO + Cr 0
a 3a
For making the mixture, the preparation: of MgO and Cr20
described above were mixed in stoichiometric proportions.
The compacting pressure ;vas 3.13 t/cm2, except of course the
experiments in B. I. t+.
1. Reaction in nitrogen atriios here.
Figure 3 shows the values obtained plotted against time.
The hardness of the compacts increased very rapidly as the
reaction proceeded; whereas the non-ignited compacts could be
crushed very easily, this was possible only critrl difficulty for
a conversion of 80% . Thu colour changed with progrc sling
conversion from chrome green to grey green.
Influence of the com-Pactinp pressure.
To ~,.ctorminc the dependence of this reaction on the
compacting pressure, compacts were pressed at different pressures
and were exposed to the same temperature for the same times.
The experimental results are plotted i Figure 4. The
values of Figure 4a, as well as curve I in Figure !h, show the
results of.elperiments in a nitrogen atmosphere, whilst curve 2
in Figure ),b shows the results of a reaction in a vacuum, carried
out by degassing for one hour before the compacts were put;hed into
the hot furnace, so as to remove the last traces of air from the
reaction chamber.
Curve 3 in Figure 4b shows the results of compacting
pressure experiments with a fresh mixture, in which sintered
Cr203 was replaced by the unsintered oxide.
The reasons for making experiments8 and 9 are explained in
the discussion of the experimental results.
Influence of oxygen.
The experiments in Figure 5 and curve 3 in Figure 6 were carried
out in an atmosphere of nitrogen in the presence of 3-4r of oxygen.
For this purpose, the commercial nitrogen was not first freed from
oxygen in a M -R tower, but merely dried in the usual way and passed
into the furnace. In the experiments shoran in curves 4 and 5 of
Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
 Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
of Figure 6. this commercial oxygen was mixed at a T -piece in the
desired proportions by volume with nitrogen from the M-R tower.
In curves 1 and 2, a current of dried oxygen and a current of
dried air were used, respectively, as gas atmosphere, the air
being taken from a compressed air generator. In the bracketed
values of curves I and 2, a loose powder mixture was used, not a
compact. In all the other experiments, the compacts were
pressed at 3.13 t/cma.
Lam. Reaction in a vacuum
The experiments in a vacuum had to be made with the pump
running, since the frequent movement of the porcelain tube
L and the wire 0 made it impossible to.maintain the vacuum when
the pump was stopped. Since leakage in of gas could occur at
the most at the movable parts at L and 0, any gas which might
perhaps find its way into the apparatus could not come into
contact with the compacts, since it would be drawn off at once at
the pump connection Q. In addition, with this arrangement, any
vapours from grease, e.g. on inserting the greased porcelain tube
into the furnace, were pumped off at once.
The results of the reactions in a vacuum are shown in
Figure 7.
With regard to the experiments in a vacuum it should also be
remarked that sintering ill a vacuum appears to proceed more slowly
than in a nitrogen atmosphere. The hardness of the compacts
after the reaction was much less than with corresponding compacts
with the same conversion, but prepared. in nitrogen. A
quantitative statement cannot be made, since there was no micro-
hardness tester available.
2 3
1. Reaction in nitrogen atmosphere.
Weighings of ZnO and Cr20a with the molar ratio I : I were
mixed in the usual way and compacts were made under a pressure of
2.5 t/cma, The results of the experiments, all of which were
made in a nitrogen atmosphere, are shown graphically in Figure 8.
The external appearance of the compacts after the reaction
were the same for colour and hardness as in the system MgO/CrO~3.
2. Influence of compacti ~r sure.
The results of the investigation into the effects of compacting
pressure in the system ZnO/Cr2O3 are shown in Figure 9, curve 2.
Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
 Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
III. Reaction CdO + Cr 0
a 3
1. Reaction in ni iro n atmospherc.
In this system also, the molar ratio was I : 1. Compacts
pressed at 1.25 t/cm2 were made from the mixture prepared in the
"Multimix". Figure 10 shows the results of the investigation
in an atmosphere of nitrogen.
As shown by the peculiar "negative" conversion values at
the commencement of the reaction, the analytical method employed
is too inaccurate for this sys tear. This is due to the fact
that the cadmium-chromium-spinal formed in the nascent state at
a low temperatures is also dissolved by hydrochloric acid, so
that a larger quantity of hydrochloric acid is used than
corresponds to the zero value, thus giving the appearance of a
negative conversion. In Figure 10, the probably real conversion
values have been extrapolated with reference to the usual forms
of the curves,
The colour of the final spinGl is chromium-green with a
yellowish tinge.
2. Influence of, compaatin --_Pressure.
The results of experiments carried, out in the usual way for
examining the effect of compacting pressure are shown in
Figure 9, curve 4.
IV. Reaction MgO + Al 0
9 3
1. Reaction in nitrogen atm. herc.
The mixture w,__.s prepared in the usual nay. The compacting
pressure was 2.5 t/cm in all cases. The A1203 preparation
employed was a sieve fraction with an aver,-,go particle size of 1+O?
In this system, the determination of the conversion was
modified by titrating the excess of hydrochloric acid directly in
the suspension after the water bath treatment, without filtering
from the solid residue. It is true that the colour change is not
so easy to recognise in this suspension, but the resulting
accuracy is perfectly adequate for the purpose of the determinations.
Figure 11 shows the results of the experiments in an
atmosphere of nitrogen.
Influence of cgmpacting russure.
As before, compacts compressed under different pressures were
heated together with the loose powder mixture (p = 0 t/cm2 ) at
the same time to the same temperature. Figure 9 shows the results.
Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
 Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
Influence of ox y en.
It was of interest to see whether in the system MgO/Al a 0 8,
which unlike the system MgO/Ox' 0 , is chemically inert to oxygen,
the oxygen content of the atmosphere exerted any effect on the
reaction velocity. For this purpose, reactions were carried out
in gas atmospheres of different oxygen contents; the results are
shown in Figure 11,
4. Influence of hydro&on _
Since the system MgO/A 1203 was the only one of those
investigated which is chemically inert to hydrogen at the
reaction temperatures employed, it was possible in this case to
ascertain whether, the reaction proceeds at a different velocity
in an atmosphere of hydrogen from that in a nitrogen atmosphere,
The results of these experiments are also shown in Figure 11.
Reaction in a vacuum
Figure 11 also shows the results of experiments in a vacuum;
the experimental conditions are riven in B,I,4.
In these investigations, a striking observation was made:
When the hot compacts were removed from the furnace and placed in
the attachment T. the originally pure white compacts became dark
blue grey in colour. To ascertain whether a different compound
had been formed, this product which surrounded the compact in a
thin layer 0.1-0.2 mm thick, was scraped off and an X-ray
photograph was made of it. No lines other than those of MgO,
A 10 and MgA1 0 were found.
a a a 1~
On ignition in the airy the blue grey colour disappeared,
and a pure white product was obtained.
Influence of carbon dioxide.
Since the starting mixture of MgO and A 10 and. also the
Al 0 sieve fraction of 4O? had been exhauster. It was necessary
2 use account of the
to use a fresh mixture of 75u was s used. On
considerably different pa- rticle size of the corundum, the
comparative value in a current of nitrogen had to be re-determined
and is shown in Figure 12.
The results of experiments made with the same mixture in an
atmosphere of carbon dioxide are also shown in Figure 12.
The ammonia gas used was taken from a bottle and was dried with
soda lime and magnesium perchlorate. Before and after the reaction,
the furnace had to be well washed out with nitrogen, since at the
temperatures employed, ammonia is almost completely decompose(a. into
nitrogen and hydrogen.
Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
 Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
Figure 12 shows the results of experiments carried out on
the mixture described in the preceding section,
8. Influence of sulphur dioxide.
The sulphur dioxide employed for this purpose was taken from
a bottle and dried with sulphuric acid and P11 05.
After the reaction, the compacts were not so strongly
sintered as in the previous reactions, Figure 12 shows the
conversion values obtained.
V. Reaction ZnO + Al$03
1. Reaction in nitroen atmosphere.
In weighing the portions for this system, an Al20a sieve
fraction of the particle size 10? was used. The compacting
pressure was 2,5 t/om2 , In this system, as in the system
MgO/Al 0 , the titration was also carried out directly in the
suspenli&n without first filtering.
Figure 13 shows the results of the experiments carried out
in an atmosphere of nitrogen.
2. Influence of eompactin ressure.
Curve I in Figure 9 shows the results of the experiments
carried out in the usual way,
34 Influence of carbon dioxide.
Figure 13 sho`-.(s the results of experiments, in which the
influence of carbon dioxide on the reaction velocity in this
system was investigated. The Las atmosphere-,employed was a
mixture of nitrogen with I CO2of pure CO
L, Influence of the particle size.
In order to investi`ate the influence of the particle size
on the reaction velocity, a number of mixtures were made with
Al 0 in different particle sizes measured under the microscope.
s
Figure 11+ shows the results of these experiments, The figure
indicates the particle size of the Al 0 for each mixture. The
zinc oxide used was the same preparation as in the preceding
experiments.
In this case, these substances could not be mixed in the
"1Multimix", because the particle sizes of the Ala03 might have
been altered by the grinding action of the machine. To avoid this,
the weighed portions, calculated in the molar proportion I : 1 were
shaken in a glass-stoppered bottle for 8-10 hours on a shaking machine.
Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
 Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
In these experiments, the gas atmosphere was nitrogen; the
compacting pressure was 2,5 t/cm2. In all the experiments, the
furnace temperature was 101$?C,
VI.. The system Cd.O/A 20a
The attempt to investigate the reaction velocity in the
system CdO/AI 0 failed, since at the temperatures necessary for
3
securing a reasonable conversion, the rate of evaporation of the
cadmium oxide was of the same order as the rate of formation of the
spinel. The experiments had to be abandoned for the time being
without result.
VII. Reaction MgO +. Fe203
In this system also, the molar ratio was I : 1; the substances
were mixed as usual and compacts were made from this mixture under
a pressure of 2.5 t/cm2 The Fe908 weighings were taken from a
.sieve fraction of the particle size 15P
Difficulties were encountered in the titration, since small
quantitic-:s of the spinel or of the iron oxide activated by the
reaction (about 1-2 mg, maximum 3 mg) were dissolvecl in the 0.1
n-hydrochloric acid on digesting the substance on the water baths
this interfered with the titration because before the change in
colour of the metkhyl red was reached, the solution became deep
yellow in colour, due to the separation of colloidal iron oxide
hydroxide with increasing pH value, this taking place to such an
extent that the change in colour could no longer be detected,
The titration was therefore modified as follows: After filtering,
2 ml of an approximately 1 molar citric acid solution was added to
the solution, ;whereby the small amounts of iron were held in
solution. In this case, in the presence of Ir'e* ions, the
solution became strongly Fellow in colour, Bromothymol blue
was therefore used as indicator, since in this solution, its
pronounced change from yellow through green to blue could be
clearly recognised.
1. Reaction in.nitrogen atmosphere.
The hardness of the compacts increased very greatly as the
reaction proceeded; for a conversion of about 30% it was already
so groat that the compacts coul& only be crushed in a steel mortar.
The colour changed substantially concurrently with the conversion from
the blue grey of the starting mixture to the red brown, colour of
the final spinel.
Figure 15 shows the results of these experiments.
Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
 Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
2,_ Influence' of compactinnoressure,
The compLLets pressed under different pressures were heated
in a nitrogen atmosphere together with the loose powder mixture.
Figure 9, curve 3 shows the results of these experiments?
Influence of oxen,
Figure 15 shows the results of the experiments in
commercial nitrogen (3,5% 02), in air and in pure ox gen at
two different temperatures,
C. Discussion of the experimental results,
Imo. Temperature dependence of the investigated reactions
in a nitrogen atmosphere,
_I Fundamental considerations on the calculation
of the reaction velocity constants with reference
to the reaction MgO + Cr 0 .
The calculation of the reaction velocity constants (R.V,C)
for determinin3 the energy of activation according to Jc,nder?s
equation
2.k.z/r! ;1 3 100 U 2
10 }
is based on the assumption of a constant particle radius.
Therefore, usinL for instance the results shows in Figure 3, we
obtain values for k which are dependent upon time. Since the
smaller grains at the conmeneement of the reaction react more
rapidly and the larger grains are left behind, the particle
size distribution varies and hence the calculated R.V,C. in
the courso of the reaction, Jagitsch (27) has already referred
to this fact in detail.
If, nevertheless, we wish to obtain a survey of the order of
magnitude of the energy of activation, the k values are plotted
against 1/T for the same reaction period. In doing this., the
time will be measured so that the preponderance of lattice
diffusion over surface diffusion can be seen from the variation
of conversion with time.
In Figure 16, the R.V.C.'a corresponding to z = 10000 sec. for
the reaction,IvlgO + Cr 0 have been plotted logarithmically against
a
1/T in the usual way,, the values of Table I being taken into
consideration. Except for t = 817?C,, the values lie fairly well
on a straight line; the discrepancy for 817?C is understandable
since such small conversion amounts are produced substantially
Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
 Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
-13-
by surface diffusion and not by lattice diffusion, and according
to Httttig (26, p, 390) the activation energies have different
values for the two kinds of diffusion.
From the straight lines of Figure 16 we obtain the following
values for the constants of'the Arrhenius equation:
k r 7,9 x 10-4 x c- 64.000/RT (cma/sec.
Reaction veloci constants and_Arrhenius constants.
Table I shows the reaction velocity constants calculated for
different systems with reference to the experimental results of
Chapter -according to the scheme given above.
The third column gives the microscopically measured mean
particle size of both substances used for the calculations; the
particle size of the MeO is in front of the oblique stroke and that
of the Me 0 is after the stroke. The constants themselves are
2 a
shown in column t~, being expressed in the unit omo/sea. X 1016.
By plotting-these" values against the reciprocal of the absolute
temperature in Figure 16, we obtain the Arrhenius straight line
of the reaction concerned, the equation for which is shown in
column 6.
As already stated, these numbers merely represent lower
limit values. Whereas in the system MgO/Al 0 Tanaka (39) found
an energy of activation of 41.1 kcal and henge?relatively good
agreement with the value of 46,C kcal found here, Bengtson and
Jagitsch (2) in the system ZnO/Al 0 obtained 98 kcal, almost
A 3
twich the value given here, although it.should be borne in mind
that in the case of the last authors, the experimental conditions
were substantially different, since they did not use a powder
mixture but two compacts pressed against each other.
Cualitatiye comparison of the
reaction t_em~eratures.
If we compare purely qualitatively the reaction temperatures
necessary for obtaining the same conversion in different systems,
it is possible even at this stage to draw some interesting
conclusions.
The maximum reactivity observed was found in the system
MgO/Fe 0 ; then come the systems CdO/Cr 0 , ZnO/Cr 0 , ZnO/A-l 0 ,
s a 2 a a s
MgO/Cr 0 e and finally MgO/A l a 0 s as system with the maximum reaction
temperatures.
These differences in reaction temperatures may be explained
partly by differences in melting point and partly by differences in
particle size. Thus, the order of melting points is Fe 0 , Al 0
A 9 a
and Cr 0 : since, however, the particle size of Al 0 is twenty
times that of Cr a 0 a , it is clear that this Al a 0 o preparation has a
Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
 Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
lower reactivity than the Cr90, and therefore comes after the
latter in the above-mentioned order. If, furthermore, we take
the vacuum reaction as standard reaction for MgO/Cr2Oa as is
proposed below (see C. II.2,a), the difference in the reaction
temperatures is still less.
In the case of thedivalent components, the order of
reactivities is analogous to that of the melting points (CdO,
ZnO, MgO). Due to the considerable differences in melting
points, the differences in particle size are not evident.
Accurate considerations concerning the necessary a -
temperatures for attaining the same conversion are not poesible
because definite statements regarding the inactive state of the
starting components cannot be made, since they were pretreated
at different a - temperatures: on the other hand, it is doubtful
to what extent the melting point values, obtained by different
methods, can be correlated.
II, Influence of gas atmosphere and vacuum.
1 Influence of oxen
(a) on the reaction
MgO + Cr?O8 .
The experimental results (Figures 5 and 6) show a
surprisingly considerable influence of the oxygen content of the
atmosphere on the reaction velocity. Curves L. and 5 of Figure 6
are most striking, where the conversion increased to more than
double for an oxygen content of fractions of a percent, It is
not surprising that this activating influence of the gas atmosphere
has a stronger effect on a loose material than on compacts
(see the bracketed measured points in curves 1 and 2).
It is very likely that this activating effect of oxygen on
the reaction MgO + Cr00a is du:e to thj fact that in the presence
of oxygen, an intermediate reaction can take place as previously
indicated by Jander (29) for the system ZnO/Cr2 0
According to this, intermediate magnesium chromate would
be formed:
2MgO + Cr20a + 3/2 0' -- 2Mg0r0. (1)
It may be assumed that this magnesium chromate is not stable
at the temperatures employed and is immediately decomposed
according to
2MgCrO -, MgO + MgOrs04 + 3/2 02 (2)
Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
 Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
In the first place, this assumption is supported by the
fact that the activation energy of this reaction in the presence
of 3-l~0 0 is only about 1 13 of the value in a pure nitrogen
atmosphere (see Table I : 19900 cal).
Further experimental support for this assumption is provided
by the fact that small amounts of chromate ions could be
detected in the filtrates of these experiments by means of the
diphenyl-carbazide reaction.
In order to lend further support to this reaction
mechanism, the decomposition temperature of magnesium chromate
had First to be determined and it had to be proved that its
decomposition takes place according to equation (2), since no
information on this point could be found in the literature.
MgOr04 was prepared by introducing MgO in small portions
into a solution of 125 g Cr08 p.a. in 250 ml of water until
the pH value was 6-7. This solution, filtered from slight
impurities was evaporated under reduced pressure at 60?C to a
dark red, syrupy liquid.
On cooling to 2 - 3?C, a thick, yellowish red crystalline
mass separated out which was filtered off and washed with
absolute alobhol and ether. The resulting product consisted of
very beautiful lemon yellow crystals.
Analysis showed a CrO content of 43.95`% for an Mg content
of 9.20jo (theoretically, for this Mg content the CrO content
should be 43.86;). This composition corresponds with
satisfactory accuracy to the formula IVIgCrO1t~~.7H~0. which according
to Hill (23) should crystallise out below .~16?C.
For determining the decomposition temperature, weighed
portions of 2-3 g of this product were heated in a platinum
crucible in the crucible furnace or in a porcelain boat in a
tube furnace and the loss in weight at different temperatures
was determined,
On heating in air, decomposition commences at 350?C
although very slowly, a deepening of the colour towards
orange being observable. At 575?C dissociation becomes more
pronounced (colour commences to change towards green) and is
practically complete at 620?C,- The entire loss of weight is
then 16.71 or 17.04X' (theoretical loss of weight according to
equation (2) is 15,86;.).
In a current of nitrogen, decomposition is already
quantitative at 6115?C, the loss of weight is 16.95%4.
In a current of oxygen, dissociation is already complete at
630?C; the orange yellow colour persists, however, to 625?C,,
and the characteristic change of colour to grey-green the colour
of spinel occurs only then. The total loss of weight is then
17,06%. Even at 575?C the weigh ble decomposition is very
considerable.
Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
 Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
-16-
If the decomposition of MgCrO occurs according to
equation (2) the MgO content of th~ ignited residue should be
17.33"; . This was found to be 17,17,,.
Thus, at temperatures of 600-620?C, magnesium chromate
loses oxygen and decomposes into a mixture of MgO and
magnesium chromite.
An attempt to obtain MgCrO4 in accordance with equation
(1) below its dissociation temperature, as Vasenin (1+0) and
Ford (5-7) were able to do in the system CaO/Cr2Oa/Oa was
unsuccessful, nor was it possible, in accordance with the
reverse equation (2) to prepare magnesium chromate from
magnesium chromite, MgO and oxygen. The chromate content of
the products which had been heated for 12 hours in a current of
oxygen at 550?C was not more than 0.1 - 0,2% .
It is, of course, not impossible that other compounds
in the system MgO/Cr 2 0 a /0 a , perhaps like those which Ford
(5-7) found in his investigations in the system CaO/CrO/0~
or other valency stages of chromium, such as recently found by
Scholder (35) exercise this activating function on the
reaction in the presence of oxygen.
(b) on the reaction MgO + .Al 4 0 3
The values of Figure 11 fail to reveal any definite
influence of oxygen on the reaction velocity; the observed
deviations lie within the reproducibility of a reaction.
The rate of formation of MgAl 0 is accordingly unaffected
a a
by the oxygen of the atmosphere.
(c) on the reaction MgO + Fe-O~
The values of Figure 15 show a distinct decrease in the
reaction velocity due to the presence of oxygen, as compared
with a pure nitrogen atmosphere. The extent of the
inhibiting effect of the oxygen at 827?C is independent of the
oxygen concentration, but at 865?C the inhibiting effect is more
pronounced at higher oxygen contents.
An attempt to explain this fact led to the following
consideration:- The equilibrium
6Fe203 L L , 4Fes02 + O
is shifted in favour of Fe20a in the presence of oxygen.
Simon and Schmidt (36) give 0.5 mm Hg as the equilibrium
.pressure of this system at 1150?C and 1.0 mm Hg for 1200?C;
other figures in the literature arc round about these values,
Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
 Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
..17
The reaction temperatures employed are indeed much lower
than these temperatures of a measurable dissociation, so that
any appreciable thermal reduction can scarcely occur. It is,
however, quite conceivable that a certain small number of high
energy lattice particles split off oxygen in the surface in
quantities which cannot be detected analytically and may be
transformed into Fe 04.
Since, on the other hand, the conversion Fe20a -s Feg04 is
accompanied by an extensive lattice transformation, and on the
other hand the active molecules, which as regards energy are
in a position for this, are situated at the important centres
of the surface for reactivity, the surface of the n FeO
s
preparation, in the absence of oxygen, may be thermally "reduced"
and the Fe O may thereby be activated. In the presence of
2 3
oxygen, this surface dissociation will be repressed and the
reaction velocity thereby retarded.
According to Hlittig (25), changes, in modification may become
notic.able, even 100-200?C below the actual conversion point,
through changes in the surface, and Anderson (1) also found such
"primary" reactions or "surface" processes in the case of
heterogeneous reactions.
2. Reactions in a vacuum.
(b) Reaction MgO + Cr203
Figure 7 shows that the reaction MgO + Cra0a proceeds much
more slowly in a vacuum than in nitrogen.
It is not ruled out that traces of oxygen still possessed by
the nitrogen purified by the M.R. Tower, and which according to
the literature (32) are less than 10"47 bring about this increased
conversion in nitrogen as compared with the reaction in a vacuum,
and therefore the vacuum reaction must be regarded as the actual
"standard" reaction, i.e. it proceeds without the influence of any
foreign gases.
(b) Reaction TvfgO + Al A 0
As shown by Figure 11, marked differences can be seen from the
reaction in nitrogen, especially at the commencement, but these
disappear in time. After 10 seconds, conversion is practically
the same in both reactions.
Thus, in the system MgO/Al08, the reaction in a vacuum appears
to proceed'with the same velocity as in nitrogen. The inhibiting
or promoting influences on the reaction velocity mentioned in the
literature could not be confirmed for the systems investigated
here and presumably are ascribable to the special features of the
systems investigated,
Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
 Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
Influence of hydrogen on the reaction MgO + Al 0
It will readily be seen from the results of Figure 11 that
in principle the same applies here as for the experiments in a
vacuum. Here again marked differences are to be observed,
especially at the commencement of the reaction, but cannot be
regarded definitely as real. Such irregularities occur quite
frequently and are ascribable to the generally difficult
reproducibility of reactions in the solid state.
In the reaction MgO + Al R 0 3 a definite influence cannot be
found in hydrogen as compared with the nitrogen reaction, nor
was such influence expected.
4.. Influence of carbon dioxide.
Although, as shown by Figure 12 carbon dioxide has no
appreciable influence on the reaction velocity in the system
MgO/Al 2 0 $ it does have in the system ZnO/Al s 0 a .
If 1% 00 was added to the usual nitrogen atmosphere, the
a
reaction velocity was practically the same as in a pure
nitrogen atmosphere; in a pure CO s atmosphere, however, a
definitely higher reaction velocity can be observed.
This difference in behaviour is difficult to explain,
since the increase in reactivity be the creation of fresh
surfaces from an equilibrium situated on the side of MeC for a
possible intermediate reaction
Me0 + CO2 McCOs
ought to be applicable for ZnO and for MgO.
Evidently, such loosening of the surface for the system
MgO/AIROa is not sufficient to influence the greater inertia of
this system, to which reference has already been made, especially
since, in this case, the A 2 ,20 $ preparation eziployed had a
considerably particle size.
5. Influence of ammonia on the reaction MgO + AI -O
As shown by the values of Figure 12 an ammonia atmosphere
does not exert any visible influence on the reaction MgO + Al 0 .
D 3
This means that the establishment of the equilibrium
2NH = N + 3H
$ i 9 a
on the surface of the reaction mixture is without effect on the
reaction velocity in this system.
Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
 Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
6. Influence of sulphur dioxide.
The values of Figure 12 show a definite reaction promoting
effect of sulphur dioxide on the reaction g0 + Ala 0
There was the possibility that considerable amounts of
MgSO a were formed in the compacts, so that the MgO employed was
no longer available for titration. Although this sulphite was
very probably decomposed on digesting with hydrochloric acid., it
is not certain whether the liberated SO a was completely expelled
on the water bath or consumed additional hydroxide in titration.
Since, however, the SO a content of a compact after the reaction
was less than 1%, the strikingly high conversion values do not
appear to have been produced by combined SO but are real.
This phenomenon was not investigated further.
III. Influence of contact possibilities.
1. Influence of compacting pressure,
If we compare the compacting pressure curves shown in
Figures L. and 9, the shape of the curves in the various systems
is very different; certain uniform characteristic effects may,
however, be observed and they appear to have a generally
valid significance.
Considering first of all the compacting ressure
dependency in the system MgO/Cr 2 0 3 (Figure L),which was
investigated more closely, it is found that the conversion at
first remains constant with increasing compacting pressure
between a pressure of 0 and 3 t/cm2, whence it increases
substantially linearly to about 6.5 t/cm2; from that point
there is only a slight increase until the maximum compacting
pressure investigated.
The fact that the conversion remains constant up to a
certain pressure and then suddenly increases permits of two
possible interpretations:
(a) As has already been shown, this reaction may be
increased considerably by traces of oxygen. There is now the
possibility that the air enclosed in the compacts during the
pressing operation, from a certain compacting pressure onwards,
is no longer displaced from the compacts in washing out the
reaction tube with nitrogen; the oxygen thus retained in the
compacts would then be responsible for the increased conversion.
(b) From the known compacting pressure onward, a destruction
of the grains of the substance occurs, thereby providing fresh
surfaces of fission and fresh possibilities of reaction, causing
the increased conversion.
Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
 Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
As shown by the results of an experiment carried out in
a vacuum with the pump running (see Figure 2+b, curve 2), the
same compacting pressure dependency occurs in a vacuum as in
nitrogen. The first of these two possibilities is thus
eliminated.
To be able to support the second possible interpretation,
it must be remembered that the hardness of the grains plays a
considerable part. This hardness depends largely upon the nature
of the pre-treatment and hence upon the sintering temperature.
For this reason, it was of interest to repeat the experiments
with a mixture of MgO and unsintered Cr203.
As shorn by curve 3 in Figure L.b, there are two increases
in the conversion by increased compacting pressure, at 0.75
and at 4 t/cm2, However, there is definitely no increase at
3 t/cm2 as in the reactions with sintered chromic oxide.
On the assumption that the increase in the conversion due
to increased compacting pressure results from the crushing of
the Or203 particles, the unsintered Cr203 appears to contain two
groups of particles of different hardness, so that at a pressure
of 0.75 or 4 t/em$ fresh possibilities of contact are provided
which cause the increase in the conversion.
To avoid errors, it must be emphasised at this point that
by "hardness" of the particles or grains we do not mean the
hardness of the substance in the mineralogical sense - because
this cannot of course be affected by sintering, but the hardness
of a sintered grain, i.e. the mechanical ability to support stress
possessed by the contact surfaces of the single crystals of which
the particle is composed.
This effect that the conversion increases from a certain
pressure due to the breaking up of the grains of substance,
besides occurring in the system MgO/Cr203 is also very
pronounced in the system MgO/Fe203 at a pressure of 5 t/crn2
(see Figure 9, curve 3). In this latter system, apart from
the "fracture effect", a second effect has been found responsible
for the increase, in the conversion: Between 0 and 3 t/cm` a
considerable increase in conversion can be observed. It
cannot be assumed that this slight pressure can already exercise
a grain-crushing effect; on the contrary, in this case the
pressure causes the particles to approach each other, that is to
say, it results in a better contact of the grains, and so will be
the cause of the increased conversion.
This effect will preferably occur at low pressures.
In the dependence upon compacting pressure in the system
ZnO/Cr 0 , two different effects can clearly be observed. In
additi&n3to the increase in conversion caused by the compacting
Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
 Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
pressure, the drop in the conversion curve between the compacting
pressure Ot/om2 and the compact with the lowest compacting.
.pressure measured is remarkable. This confirms the phenomenon
found by Fricke (1+) and Forestier (10) and the reaction in the
compact, particularly at low compacting pressures, may proceed
more slowly in the compact than in the loose powder. Fricke
explains this by the fact that, in the loose material, the points
of growth can adjust themselves to each other more readily than
in the pressed material. Perhaps this phenomenon may be
expained better by stating that certain active corners and edges,
which are situated in the surface of the grains and along which
diffusion to the reaction partner can take place, become blocked
in the compacting process so that they are not available for
diffusion.
Apart from the system ZnO/Cr203, this phenomenon is also found
to be very pronounced in the system ZnC/AlO and less pronounced
in the systems CdO/Cr 2 0 3 and MgO/Al 2 0 3 (see.Figure 9). Apart from
such "blocking", the three last-mentioned systems do not appear
to be affected in their reactivity by a variation in the compacting
pressure.
Accordingly, three different effects are to be distinguished
in the influence of the compacting pressure on reactions in the
solid state, which probably overlap each other in their
dependence upon the nature and structure of the grains.
(a) An increase in conversion, due to the purely geometrical
approach of the particles.
(b) An increase in conversion, due to the crushing of
individual substance grains at certain compacting pressures..
(c) A reduction in conversion by the blocking of possible
favourable means of diffusion in the surface.
For a rational interpretation of the compacting pressure
curves it would probably first be necessary to have statistically
adequate' experimental material, for which a satisfactory
interpretation would be possible by systematically varying the
experimental conditions and suitably selecting the components.
2. Influence of the particle size.
This influence was investigated only in the system ZnO/Al0
As Figure 11+ shows, the conversion obtained for all the particle
sizes of Al 0 employed is the same within the limits of error,
2 3
although the latter vary by the factor of three. Any possible
conclusion can therefore at present only refer to this special case.
However, this peculiar result at first appears. to contradict all
existing notions concerning the dependence of solid state reactions
Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
 Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
upon the particle size (.27, 28 inter alia). It might :possibly
be explained according to Tanaka (39) by the fact that the Al20
migrates on the ZnO or, in the case of the formation of a reaction
layer, diffuses into the latter to the ZnO.
According to his experimental results, Tanaka considers this
mechanism to be very likely. He believes that above all in
reactions, whose character is strongly determined by surface
diffusion, the particle size of the diffusing component is only
of slight significance, whilst the particle size of the component
on which the reaction product grows, affects the reaction velocity
in the sense of Jander's formula (28).
Such an assumption of the diffusion of Al 2 0 a to.ZnO, however,
is in direct contradiction with the results of Bengtson and
Jagitsch (2), who found that the direction of migration was
definitely in the opposite direction for the system ZnO/A1a03.
Apart from the fact that all thexmal data support the results
of Bengtson and Jagitsch, the considerable difference in the
particle sizes should be borne in mind in the present case. It
may act upon the conversion like a variation in the mixing
proportions of 1:1, depending upon the excess of the component
whose direction of migration determines the reaction.
In conclusion, the influence of the mixing method employed
on the conversion of the reactions herein described will be
considered.
If we compare the values of Figure 13 with those of Figure
14, we find that in the former, the conversion values are
substantially higher (for example about twice as high for 1000?C.
The only difference in treatment of the two starting mixtures
was that the preparation of the mixtures for Figure 13 took place
in the "Multimix" and for Figure 4 on the shaking machine.
The objection that the higher conversion in Figure 13 was
caused by a grain-crushing action of the Multimix is invalid,
because the results of the above section have clearly shown
that the particle size of the Al 0 is without influence on the
s 3
reaction velocity; since, on the other hand, the particle size
of the ZnO is only about 2 - 5% of that of the corundum, it is
unlikely that its particle size has been materially altered by
the Multimix treatment.
The reason for this difference probably resides in the fact
that the individual ZnO grains readily agglomerate to form very
small, loose spheres which persist in the mixing on the shaking
machine, so that a homogeneous mixture is not obtained. These
spheres, however, are not stable to the vigorous mixing action
of the."Multimix" but are broken up into their constituents.
Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
 Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
-23-
For this reason, therefore, a much more homogeneous mixture
can be produced in the "Multimix" than by mixing of moderate
duration on the shaking machine, since it cannot be assumed that
a mixture which after 10 hours is still at least 50% from being
an ideal mixture, will have approached this degree of mixing to
any considerable extent after say ten times as long. This shows
the extent to which the results of solid state reactions depend
upon the working conditions.
IV. Final remark.
The results obtained in the present work do not serve to
simplify the by no means very simple picture of the reactivity of
solid. It is shown that the course of solid state reactions
may be influenced in very different ways. It is at present not
possible satisfactorily to co-ordinate the observed facts with
reasonable notions of the mechanisms of such reactions.
The possible interpretations proposed in the paper are to be
supported by enlarging the experimental material .
Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
 Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
E. Summary.
1. It is intended to provide a contribution to the .problem
of the influence of foreign gases and contact possibilities on
the reactivity of solid substances.
2. An apparatus is described, in which the course of
solid state reactions with time can be followed in definite
gas atmospheres.
3. A simple analytical method is given for determining the
conversion in spinel reactions.
5. The temperature dependence of the reactions MgO + Cr a 0 s,
ZnO+Cr0, CdO + Cr 2 0 3 , MgO + Al 2 0 3 , ZnO + Al 2 0 3 and Mg0+Fe 2 0
2 3 3
has been investigated and the corresponding constants of the
Arrhenius equation given.
5. In the action of the compacting pressure on the reaction
velocity, three different effects are to be distinguished; in
certain circumstances, these effects overlap; they were found
definitely in the following systems:
(a) MgO/Fe203
(b) MgO/Fe 2 0 3 and MgO/Cr 2 0 3 ,
(c) NnO/Cr 2 03 and ZnO/A1203.
The systems MgO/A1 a 0 3 and CdO/Cr a 0 3 show little or no
dependence of the conversion upon compacting pressure.
6. Whereas in the system MgO/Al 2 0 3 the particle size of the
Al 0 has a distinct influence, in the system ZnO/A l 0 a
2 3 2 3
variation of the particle size of the Al 2 0 S by the factor 3 has no
effect on the conversion obtained.
7. If the starting mixture is mixed in one case on the shaking
machine and in the other case in the Multimix, there is a difference
in the degree of mixing which, in the case of the latter method of
mixing and in the system here investigated, results in a reactivity
which is almost twice as great.
8. Oxygen promotes the reaction velocity in the system
MgO/Cr203 very greatly, so that oxygen contents of a few tenths of
a percent can double the conversion. There probably occurs here
an intermediate reaction of the type.
2 MgO + Cr203 + 3/2 02 -* 2MgCrO 4
2 MgCrO4 -* MgO + MgCr204 + 3/2 02
To confirm this, the decomposition temperature of MgCrO4 was
determined; it is 610 ?10?C. In addition, it was shown by
analysis of the solid decomposition product that the decomposition of
Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
 Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
MgCrO4 takes place in accordance with equation (2)
In the absence of intermediate reactions which are
influenced by C 2 , the latter was fpund not to have any effect.
Thus, for example, in the system M\lgO/Al a 0 e the reaction in
oxygen proceeds at the same velocity as in nitrogen. On the
other hand, oxygen exercises a definitely inhibiting effect on
the reaction MgO + Fe a 0 3 as compared with the same reaction in
nitrogen. This is explained by the fact that in a nitrogen
atmosphere, a surface thermal dissociation of Fe303 to Fe30
may occur substantially below the actual dissociation
-temperature, whereby the Fe 2 0 3 is reactivated. This thermal
dissociation is repressed by the presence of oxygen and the
activating action of the nitrogen is cancelled.
9. In the temperature range investigated, hydrogen has
no appreciable influence on the reaction MgO + Al
a 0 3
10. The same also applies for equilibrium mixtures of
gases. In the system MgO/A7 $ 0 3 , ammonia gas or the presence
of the components II 2 and N 2 from the equilibrium 2NH N a + 3H
3 a
is without influence on the reaction velocity.
11. It was not possible to detect any influence by CO on the
reaction velocity in the system MgO/Al a 0 3 , whilst the reaction
ZnO + .Al- 2 0 3 is distinctly accelerated by CO 2 .
12. Gaseous SO increases the conversion in the system
a
1V1gO/A l 0 to a considerable extent, an intermediate reaction
2 3
participating in certain circumstances.
13. The reaction MgO + Cr203 takes place in a vacuum
(10-1 mm I-Ig with pump running) much more slowly than in nitrogen,
it being not impossible that the traces of oxygen still present
in the purified nitrogen are the cause of this increase in the
conversion.
The reaction MgO + Al203 has substantially the same
velocity in a vacuum as in nitrogen.
Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
 Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
REFERENCES
(3)
(4)
J.S. Anderson, Discuss. Faraday Soc., 1J (1948) 163.
B. Bengtson and R. Jagitsch. Ark. Kem. Mineralog. Geol. 2"
No. 18 (1947).
F.V. Bischoff. Z. anorg. allg. Chem. 250 (1942) 10.
R..,Fricke and F. Blaschke. Z. anorg. allg. Chem, 251 (1943)
396.
(5) W.F.
4.8 Ford, W.J. Rees and J. White, Trans,:Brit. Ceram. Soc.,
(1949) 291.
(6) W.F. Ford and W.J. Rees. Trans. Brit. Ceram. Soc. 47 (1948)
207.
(7)
W.F. Ford and J. White. Trans. Brit. Ceram. Soo. 488 (1949)
417.
(8) H. Forestier and J.P. Kiehl. C.R. Acad. Sci., 229 (1949) 47.
(9) H, Forestier and J.P. Kiehl. C.R. Acad. Sci., 225 (1949) 197.
(10) H. Forestier, C. Haasser and J.L. Longuet. Bull. Soc. chim.
France, 116 (1949) D 146.
(11) H. Forestier and J.P. Kiehl. C.R. Acad. Sci., 230.09502288.
(12) H. Forestier and J.P. Kiehl. C.R. Acad. 232' (1951 )1664.
(13) Gmelin, Handbuch der anorg. Chemie, 2L Part B (1939) 47.
(14) Gmelin, Handbuch der anorg. Chemie, a Part B (1924) 71.
(15) I.A. Hedvall and K. Olsson. A. anorg. allg Chem.,243
(1939) 237.
(16) I.A. Hedvall and 0. Runchagen. Z. Naturforsch. 28 (1940) 4.29.
(17) I.A. Hedvall, Ing. Vetensk. Akad. Hand1?(1942) 48.
(18) I . A. Hedvall. Glas techn. Ber., 20 (1942) 34.
(19) I.A. Hedvall and T. Gunther. Z. anorg. allg. Chem,,2 1 (1943)
305.
(20) I.A. Hedvall and A. Lundberg. Ark. Kern, Mineralog? Geol., 17 A.
No. 12 (194)+).
(21) I.A. Hedvall, Silicates Ind, 16 (1951) 157.
Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
 Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
REFERENCES (Contd).
(22) I.A. Hedvall. EinfDhrung in die Festktirperchemie, Brunswick,
1952.
(23) A. Hill, G. Soth and J. Ricci. J. Amer, Chem. Soc. 62 (1940)
2131-
(24) G.F. Huttig and E. Ktirschner. Z. Kolloidchem 81 (1937) 40.
(25) G.F. HUttig. Z. angew. Chem. (1940) 35-.51 (26) G.F. Htittig in : Schwab, Handbuch der Katalyse, Springer,
Vienna, 1943, Vol. 5-
(27) R. Jagitsch. Chinia (Zurich) 1 (1947) 105-
(28) W. Jander. Z. anorg. allg. Chem., 1j (1927) 1-
(29) W. Jander and K.T. Weitendorf. Z. Elektrochem. angew. physik
Chem. 1+1 (1935) 435.
(30) J.P. Kiehl. C.R. Acad. Sc. 232 (1951) 1666.
(31) J.P. Kiehl. C.R. Acad. Sc. (1952) 943.
(32) F.R. Meyer and G. Runge. Z. angew. Chem. 52. (1939) 637-
(33) M.S. Roberts and H.E. Mervin. Amer. J. So., (5) 21 (1931)
145-
(34) R. Ruer and M. Nakamoto. Rec. Trav. chim. Pays-Bas. 1?2
(1923) 675.
(35) R. Scholder. Z. Elektrochem. Ber. Bunsenges. physik,
Chem., 56 (1952) 879.
(36) A. Simon and Th. Schmidt. Z. Kolloidchem., Erg.-Bd., 36
(1925) 77.
(37) P. Stahl. C.R. Acad. Sc., (1951) 1669.
(38) Y. Tanaka. Bull. Chem. Soc. Japan, 16 (1941) 428.
(39) Y. Tanaka. Bull. Chem. Soc. Japan, 11 (1942) 229.
(4o) F.I. Vasenin. C. A. 1?2 (1948) 8594.
Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
 Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
TABLE 1.
systems dealt with in the present paper.
System Conditions Particle
Size
Reaction
velocity t(?C)
constant
Arrhenius
equation:
k
MgO/CrO
1:1,N
2/2
4,2
1 817
a 3
2
15,7
915
74,0
997
592
1095
1970
1195
7.9.10"'4?e-64000/RT,
MgO/Cr
0
1 :1,N
+ 3,5% 0
2/2
2L6
817
2
3
2
:;
500
903
915
} 997
2,410"10? e-19900/RT
Mg0/Cr 03
1:1 Vacuum
2/2
4,2
915
46
317
995
1095
_3? ..69000/RT
3.7 10 e
0
Zn0/Cr
1:1,N
2,6/2
21
820
3
~
3
162
905
{
426
950
818
1390 90
030
1030
56000/RT
3,5.10-4..e"
CdO/Cr 0
1:1,N
6/2
23,4
783
2 3
2
i
130
865
290
1140
910
978
1,2.10"4?e-52000/RT
Y,gO/Al 0
a a
1:10N
2/40
53
918
780
1018
1 440
1048
3310
7860
1125
1217
4, 2.10 ? e -46000/RT
Zn0/A1~03
1:1,N
2,6/40
110
835
2860
900
13150
28400
985
1055
9.1.10-6?e-45400/RT
0
MgO/Fe
1:1 .N
2/15
45
783
a
3
770
827
1920
865
8250
13200
927
1020
1.3.10-1. e-61400/RT
Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
 Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
(In these Figures. U ? conversion and 2 = TIme)
Fig.3: Mg0 + Cr903 I:1;
N2 atmosphere.
3.13 t/v2;
Fig.4: MgO + Cr203.
Compacting pressure dependence.
(a) Reaction In N?
(b) --: _ N
+ - . Vlcuua
o Vn?lntered Cr?03
sic .tin -" n w +faT
7 1,W
-VT
Fig. 5: MgO + Cr203 1:1; 3.13 t/cs2
N2 + 3.596 02
Fig.6: MgO + Cr203 1:1; 3.13 t/as2;
t - 915aC; 02 influence.
Fig.?: Mgo + Cr203 1:1;
3.13 t/v2; Vacuum.
Fig-8: Zn0 + Cr2O3 1:1; 2.5 t/cs2;
L
N2 atmosphere.
Fig.9: ZnO + Cr203, CdO + Cr203, MgO + A1203,
ZnO + A1203, MgO + Fe203;
Compacting pressure dependence.
1.
Z __ ? ZnO +
Al203
1.
-a-- a ZnO +
Cr?03
3._0_ . Mdo +
Fe?03
- CdO +
Cr
0
4.
a
2
3
S.- + s MgO +
Al?03
Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
 Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3
Fig.10: CdO + Cr203 1:1; 1.25 t/o2;
N2 atmosphere
- - - - Probably real values.
Fig-11: MgO + A1203 1:1: 2.5
O = NZ + 3.3%
03. 1018'C
D a
H
.
1010C
Aa N3 + 30%'
03, Iolaec
+ -
3
H3.
111 Sec
o a 03
1018eC
A
H?.
11080C
r = vacuum,
1018?C
jr =
H3,
1212eC
A = racuua,
1105?C
+' :o.,. -
F19.13: Zoo + A1203 1:1; 2.5 t/cat,
N2 atmosphere and W2 influence
? - - reaction in N 2
x a reantl.,., r., At . ... - ___.
o - reaction in CO
2' 910eC.
------amr
t/cm2;
6~: Avr
v . ' yin a as
Fig.15: Mg0 + Fe203 1:1; 2.5 t/cs2;
N2 and 02 influence
reaction in N2 ataoephero
Q- reaction in N2 + 3.8% 03. 837eC
? - reaction in N7 + 30% 03. 8270C
6- reaction In O2 atmosphere, 82Y*C
? - reaction in N3 + 3.3% 03, 8630C
? - reaction in N'3 + 20% 03. 888?C
Fig-12: Mg0 + A1203 1:1; 2.5 t/cm2;
NH3 and S32
x = CO2. 1068?C 13 = NH3. Io3SoC
+ a CO 2- 1116=C 6 = sot, 1016eC
- reaction in N.
= Ho 3. 1060eC
F1g.14: ZOO + A1203 1:1: 2.5 t/cat; N2
Dependence upon particle size.
Oz = A1303 particle site
53 g
+ - A1303 particle site
73,U
ot]= A1303 particle miss
135/2
0 - A1203 particle also
155,U
Flg.18: Arrhenius straight lines for the
treated systems.
1180 + Cr303.
N3
2.--e-
N40 + Cr303,
N2 + 3.3%
3.-+--
MOO + Cr303.
vacuum
4.-o-
!e0 + Cr303.
N3
CdO + Cr303.
N3
MOO + A1303.
N3
=e0 + A1a03.
N3
M
o
a
+ 1.303,
N3
Approved For Release 2007/10/23: CIA-RDP78-04861A000400030012-3