By Howard F. McMurdie and Esther Golovato I. Introduction

U. S. Department of CommerceNational Bureau of Standards

Research Paper RP1941Volume 41, December 1948

Part of the Journal of -Research of the National Bureau of Standards

Study of the Modifications of Manganese DioxideBy Howard F. McMurdie and Esther Golovato

Past work on the modifications of manganese dioxide of interest in dry-cell manufactureis reviewed. New X-ray data, at both room and elevated temperatures, combined withdifferential heating curves lead to the conclusion that five types of manganese dioxide exist:(1) well-crystallized pyrolusite; (2) gamma manganese dioxide, a poorly crystallized pyrolu-site; (3) ramsdellite; (4) cryptomelane, a form containing essential potassium or sodium;and (5) delta manganese dioxide, believed to be a poorly crystallized cryptomelane. Thehigh-temperature X-ray diffraction data indicated the phase changes that cause the heating-curve effects. A new crystal form of manganosic oxide (M113O4), stable above 1,170° C, w^sfound to be cubic of spinel structure. Fineness determinations by both the nitrogen adsorp-tion and the X-ray line broadening methods were made on selected samples.

I. IntroductionDuring the years 1940-46 there was increased

research on (Jry cells. This was stimulated byincreased demand for the cells as well as new usesfor them, combined with certain shortages of rawmaterials. This work disclosed among otherthings that manganese dioxide is not a simplecompound with constant properties, and that itsvalue as a depolarizer depends on properties otherthan merely purity. At that time a paper waswritten at this Bureau [1] x giving some prelim-inary findings. Since then, additional work hasbeen don'e both here and elsewhere [2] on man-ganese dioxide. The present paper is an attemptto evaluate the work done here and that reportedby others. The work has not been confined todry-cell technology, but was aimed toward abetter general understanding of the oxide, itsvarious forms, transformations, and means ofidentification. This information, it is hoped, willbe of interest in mineralogy and crystallographyas well as in electrochemistry.

II. Apparatus and MethodsIn the present study, X-ray patterns were made

on the North American Philips Geiger CounterX-ray Spectrometer by using unfiltered FeK radi-ation. The patterns were automatically recordedwith a counter movement of 1° 0/min. With

i Figures in brackets indicate the literature references at the end of thispaper.

this equipment a flat specimen is used, and nospecial techniques were employed to prevent pre-ferred orientation. It is realized that in a fewcases this may have resulted in relative intensitiesthat differ from those in other reports. Thisequipment in its commercial form is not capableof recording the diffraction effects at angles greaterthan 45° 0; thus; the back reflection lines aremissed. Some conclusions are based in part onprevious studies with the use of photographicmethods.

The X-ray diffraction patterns at elevated tem-peratures were made in the apparatus describedby Van Valkenburg and McMurdie [3]. Withthis, patterns could be obtained at various ele-vated temperatures with no intermediate cooling.In many measurements at high temperature, onlythe section of the*pattern corresponding to anglesless than. 25° 0 was scanned, as this was found tobe sufficient to identify the phase present.

Differential heating curves on MnC>2 were madeon 2-g samples by the method outlined by Speiland others [4]. A12O3 (corundum) was used as aninert body. The temperature of the MnO2, andthe difference in temperature between the MnO2

and the inert body were automatically recordedas the temperature was raised at a rate of about8° C/min. The equipment used was described byNewman and Wells [5].

Electronmicrographs were made with the RCAmodel EMU microscope.

Modifications of Manganese Dioxide 589

III. The Crystal Modifications of MnO2

Although there is more general agreement on theforms of MnO2 now than existed several years ago,there still is a certain amount of disagreement onthe relations between the various forms. For ex-ample, all agree that gamma MnO2 is a poorlycrystallized material, but Cole and his coworkers[2] claim that it is closely related to ramsdellite,whereas McMurdie [1] has considered it related topyrolusite. As a first step toward settling thesequestions, it must be agreed as to what lines inthe X-ray diffraction pattern are essential to de-lineate a particular modification.

There is general agreement on the well-crystal-lized tetragonal form of MnO2, pyrolusite. Thiswas made artificially by Ferrari [6] and the crystalstructure worked out. It was found to be of rutilestructure with a=4.44 A and c=2.89 A. Thelines for angles less than 40° 0, which are presentin an X-ray pattern, using Fe radiation, are givenin table 1. It will be seen that some of the linesreported by Fleischer and Richmond [7] are

zLJ

<

liJ U

1,1 l 1 i I . I

I I M i l , i

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40DEGREES 9

FIGURE 1. X-ray diffraction patterns of various modifica-tions of M11O2.

a, No. 45 pyrolusite; b, No. 42 ramsdellite plus pyrolusite; c, No. C16 gammaM11O2; d, No. 4o cryptomelane; e, No. Cl delta M11O2.

inzLJ

>

<

I I I I I

I I I 10 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40

DEGREES 6

FIGURE 2. X-ray diffraction patterns of various samples ofMnO2.

a, No. C15, pyrolusite; b, No. C9 gamma M11O2 plus cryptomelane; c, No.Cll delta M11O2, d, No. 46 gamma M11O2; e, No. C8 gamma M11O2.

missing. These extra lines are not compatiblewith the structure given by Ferrari and are notfound here in the purest samples. They aredoubtless due to impurities. In figure 1, a,2 thepattern of pyrolusite obtained from a sampleof natural ore from Egypt is represented. The110 and the 220 lines are possibly unduly strongbecause of preferred orientation in the flat holder.Figure 2, a, shows the pattern obtained from arti-ficial material made by heating Mn(NO3)2. Thispattern is similar to that of the one in figure 1, a,except for the increased broadness of the lines dueto smaller crystal size.

The mineral name "polianite" was used formany years to indicate the well-crystallized form,as distinct from massive material called "pyrolu-site". It has been shown beyond question [10] byX-ray diffraction that the two are identical andthe name polianite is now discarded. It was a

2 The 0 (theta) values indicated on these figures are the angles of diffractionand are related to the d values shown in the tables by the Bragg equationd=X/2 sin 0, where X=wavelength of FeK a-radiation (1.936 A).

590 Journal of Research

TABLE 1.

Pyrolusite (No. 45)

hk l

110101200111210211220

d

A3.142.422.222.121.9851.6351.570

HI

100801050108030

~K-ray powder data

Gamma (No.

h k l

110101

111

211

d

A3.22.4

2.11

1.63

C16)

R I »

60100__

70

80

Ramsdellite T2]

d

A4.042.542.432.332.132.051.901.651.61

RI

Very strong.Strong.Medium.

Do.Medium strong.Very weak.Medium.Medium strong.

Do.

aAll lines broad.

common belief years ago that the mineral thatwas without external crystal faces (then calledpyrolusite) was superior for dry cells, because itwas hydrated. There is at present no evidence ofa hydrated form of MnO2, no other essentialelements being present.

Wad is a name that has been used to designatea massive or earthy mixture that is largely MnO2,but it often contains large percentages of H2O andBaO. It has no definite mineralogical meaningand actually consists of pyrolusite, psilomelane,cryptomelane, and other minerals [16].

A second well-crystallized form, which wasreported first by Fleischer and Richmond [7] isramsdellite. This is said to be orthorhombic, butthe unit cell dimensions have not been given.Cole and coworkers [2] reported a pattern for thismaterial that can be derived from that of Fleischerand Richmond by subtracting the lines of pyro-lusite from the latter. Of the various samplesof ramsdellite studied here, all have containedpyrolusite in varying degrees. It is believed,therefore, that the best pattern of ramsdellite isthat reported by Cole. This is reprinted in table1. The pattern of the purest ramsdellite obtainedhere is shown graphically in figure 1, b. Rams-dellite is not known to have been made artificially.

The modification of the greatest interest inbattery technology, and about which the mostconfusion has arisen is gamma (T)3 MnO2. Thiswas first named by Glemser [8] and has beennoted by many workers [9, 1]. It is found innatural ores and in artificial products and isgenerally considered a desirable form for battery

3 The terms gamma and delta used here in connection with types of MnOare not considered to refer to distinct mineral forms, but rather to denotemodifications of technical interest.

use. The patterns reported for it vary, someworkers cataloging several varieties [2]. Thepattern reported by McMurdie [1] has been criti-cized as not including a line at a spacing of about4 A (about 14°0). Although it is true that inmany cases, such as shown in figure 2, b, and figure2, d, such a line does occur with a pattern that isotherwise mainly that of gamma, in other casesit is much weaker or absent. This would seemto indicate that it is not caused by the same phaseas produces the gamma pattern and is not es-sential to it. The pattern shown in figure 1, c, isof an oxide made by pouring concentratedMn(NO3)2 solution dropwise onto a hot plate near400° C. This, except for the lack of the 4-A line,is approximately what is accepted as a gammapattern. When this is compared with the pyro-lusite pattern, it is seen that it is similar andvaries in having broader lines (indicating smallercrystal size) and the absence of the 200, 210, and220 lines. This would indicate a substance verypoorly developed perpendicular to the c axis, andof very small crystal size. The cause of the lineat 4 A, which occurs in some cases, is not known;it may be related to ramsdellite or groutite(HMnO2) reported by Gruner [11]. Both of thesephases give strong lines at about 4 A and may bepresent in poorly crystallized form with the poorlyformed pyrolusite (gamma). The fact that vari-ous differences in the gamma MnO2 pattern occuris not surprising, as the degree of crystallizationand percentage of a second phase may vary overwide limits, thus giving a considerable range ofpatterns.

It is thought very unlikely that the differences,such as noted by Cole and coworkers [2], are causedby real differences in structure; they are probablydifferent in degree of crystallization and of im-purities. The pattern, as shown in figure 1, c, isbelieved to show the minimum number of linesthat delineate this modification of MnO2.

Cryptomelane occurs as a natural mineral, hasbeen prepared (synthetically), and found to beisostructural with hollandite and coronadite [12,13]. It is the alpha MnO2 of Du Bois [18]. Theseminerals have been given the general formulaR Mn8Oi6. R is K or Na in cryptomelane, Pb incoronadite, and Ba in hollaniiite. The namecryptomelane was proposed for the alkali-bearingcompound by Richmond and Fleischer [14], itbeing the commonest of the minerals of this type.


The unit cell as determined by Ramsdell [15] wasfound to be body-centered tetragonal with a=9.82A and c=2.86 A. The pattern given in table 2and shown in figure 1, d, agrees with this cell.

TABLE 2. X-ra?/ powder data

Cryptomelane (No. 43)

hkl .

110200220130400121

d

A6.924.913.473.112. 162.40

R I

905030

1002040

Cryptomelane (No. 43)

h k l

240301141600251

d

A2.212.161.8351.641.54

R I

2025203015

Psilomelane is a name that has been used inconnection with manganese dioxide minerals, andvarious formulae have been proposed for it. Themost recent usage restricts psilomelane to ahydrous barium compound (BaMnMn8Oi6 (OH)4)[16] identical with romanechite. This mineraldoes not appear to be common among oxidesproposed for battery use. Its X-ray pattern,however, is similar to that of cryptomelane andsmall amounts of psilomelane in mixtures areeasily mistaken for cryptomelane.

Delta MnO2 is a modification, similar to gammaMnO2 in that it is poorly crystallized and thatthere is conflicting data on the essential lines.This modification was so designated by McMurdie[1] and is believed to be the same as the materialcalled amorphous by Gruner [12] and manganousmanganite by Feitnecht and Marti [17]. Cole andcoworkers [2] prepared a sample giving the twolines reported by McMurdie (at 2.41 and 1.42 A)and also a type with more lines, particularly oneat 7.13 A. These samples in many cases could bechanged to cryptomelane by boiling or heating.With the peiger-counter equipment in use here atthe moment, the patterns of various samples havebeen repeated. This apparatus does not indicatethe lines with as small interplanar spacings as wasobserved by the film method formerly used, butdoes give better delineation of lines at greaterspacings. The fewest lines found on a sample inthe 9 range covered are shown in figure 1, e.Another pattern is shown of a similar sample withadditional lines in figure 2, c. One sample ofdelta MnO2 was found occurring naturally and isshown in figure 3, d. The evidence is that delta

i l l 1 1 I

I II I

J_L .1 .0 2 4 6 8 10 12 14 16 18 20 222426.28303234363840

DEGREES 9

FIGURE 3. X-ray diffraction patterns of various samples of

MnO2.

a, No. 34 gamma M11O2 plus cryptomelane; b, No. C5 cryptomelane; c, No.C9 gamma MnO2 plus cryptomelane; d, No. 39 delta MnO2; e, No. 26pyrolusite.

bears to cryptomelane a similar relation as that ofgamma to pyrolusite in being very finely crystal-line ; and since various degrees of crystallinity canexist, the pattern will vary.

Consideration has been given to the possibilityof a hydrated form of MnO2. Such a compoundhas not been found. Weiser [20] made MnO2

samples by various wet methods, and although noX-ray data are given, from samples prepared bysimilar methods by others, it is known that thesamples must have included delta and gammaspecimens. These were dried at various tempera-tures, and he states that no evidence was foundthat hydrates exist and that the water of the finelydivided forms is adsorbed.

IV. Occurrence of the ModificationsPatterns have been made here on samples, both

artificial and natural, that hav<e been proposed foror used for battery depolarizers. The principalMn©2 modification, as determined by X-ray dif-fraction, and the source of the sample are givenin table 3. The method of manufacture and oftreatment of some commercial ores is not known.


Although it would add to the value of this reportif these data were known, it would also be of valueto have exact comparative data on the behavior ofthese ores in cells. In many cases, however, dataare available by which ores can be roughly clas-sified. In column 5 of table 3, ratings are given onthese for which some information was available.The rating from A to D in descending order ofvalue or capacity is admittedly rough, but isbelieved to be of some interest. The numberassigned to. an ore is that used throughout thereport.

TABLE 3. Samples of manganese dioxide

TABLE 3. Samples of manganese dioxide— Continued

Sample

1o

345

7g9

10

1112131415

1617181920

2122232425

26-27282930

313233 *3435

3637383940.

Source

New Brunswick- _JavaMexicoNova ScotiaUtah

ArizonaMexicoAustraliaCaliforniaMorocco

Mexico -_CubaGold CoastGreeceArkansas

MexicoPhillipinesVirginiaJavaPapua

South AmericaCubaMoroccoMontanaAfrica

CaucasiaCaliforniaIndiaMontanaBrazil

NevadaWyomingChinaHunan China __India

UtahNova Scotia

doCanadaMorocco

Pattern

Natural samples

Pyrododo

. . . . dodo

doPyro+cryptCryptGamma.Pyro

CryptPyroGamma __

doPyro

dodo

Crypt . .PyroGamma

Pyrododo

•doGamma+crypt _

PyroGammaPyro+cryptCryptPyro

dodo

. . doGamma+crypt _Pyro+crypt

CryptPyro

doDelta.. - _ -Pyro

Percent-age ofMnO2

7286818382

7878816891

78

6282

6169738888

7887878374

866684

678381

83

6988886891

Rating*

CCDCD

CDBAD

CCABC

DCBBA

CDCCA

BACB

CCCAC

BCBBD

Sample

4243

45

4647 .48

ClC2

C3C4:C5C6C7

C8C9C l lC13C14

C15

C16

Source

Lower California.-New MexicoTombstone, ArizVirginiaEgypt

Gold CoastJapanSugar Stick, Ariz_.

ChemicalElectrolytic.

No. 10 "treated"No. 12 "treated"..No. 11 "treated"No. 13 "treated"Electrolytic

Chemicaldododo

Electrolytic

Pyro fromMn(NO3)2.

Gamma fromMn(NO3)2.

Pattern

Natural samples

Pyro+cryptRam+pyroCrypt- -_

doPyro

GammaPyroCrypt

Artificial samples

DeltaGamma-

dodo

CryptGammaGamma+crypt _

GammaGamma+crypt.Delta

---. do. . .Gamma. _

Pyro

Gamma

Percent-age ofMnO 2

76

>907599

85

Rating

C

B

B

A

AABAA

A

A

In this work no attempt has been made toprepare material in various ways, but from thefindings reported here and in previous papers cer-tain conclusions can be drawn.

Pyrolusite (well crystallized) occurs commonlyin nature and can be prepared artificially byheating Mn (NO3)2; by hydrolysis of MnCl4 [2],and by heating certain cryptomelane or deltasamples [2]. Gamma samples become bettercrystallized and form pyrolusite on heating [2, &].It has been shown by several investigators [2, 7]that ramsdellite changes to pyrolusite on heatingnear 500° C.

Gamma manganese dioxide (poorly crystallizedpyrolusite) occurs to some extent in nature. Thewell-known Gold Coast ore is largely of this form,and other occurrences are listed in table 3. Thegamma manganese dioxide was prepared here byheating a saturated solution of Mn (NO3)2

quickly and is of common occurrence in electro-lytically prepared samples. Cole [2] prepared


gamma chemically by various means but wasunable to foretell when gamma and when cryp-tomelane would result. The patterns of samplesC3 and C4, table 3, indicate that the "treatment"used formed gamma from well-crystallized pyro-lusite. The treatment was done by a commercialbattery company and is not known to us. Fromdischarge data available, it is evident that thereduction of crystal size (and thus increasedsurface) resulting from the process improved theore for battery use.

Cryptomelane has been made by Cole [2] and inthis laboratory by precipitation of MnO2 fromKMnO4 solution. In some cases, the poorlycrystallized delta MnO2 resulted, and cryptomelanewas made from this by heating. Cole [2] foundthat electrolysis of solutions containing K or Pbgave a material with a cryptomelane pattern.(The ,pattern of coronadite is similar to that ofcryptomelane and may be the form resulting inthe case of Pb). Sample C5,4 with presumablythe same treatment as samples C3 and C4, gave acryptomelane pattern, just as did the untreatedmaterial. Cryptomelane, of an extremely fibrousnature, was found to result from autoclavingcertain samples of delta and gamma MnO2 [1].In nature cryptomelane is very abundant [7].

Delta MnO2, believed to be poorly crystallizedcryptomelane is found rarely in nature, only onesuch sample being available here (sample 39, fig.3, d), but it is formed from KMnO4 solution byprecipitation with HCl. It is found as a majorconstituent in many chemically produced oxides.

Ramsdellite is known to occur with certaintyonly at Lake Valley, N. Mex. Here it is inti-mately related to the pyrolusite. It is not knownto have been made artificially.

V. Particle Size and ShapeBecause the action of MnO2 in dry cells is

largely a surface action, the degree of fineness isimportant. The gamma and delta MnO2 arefinely divided, as shown by the line broadening inthe diffraction patterns. This fine crystal sizeand high surface area are also shown by theelectron microscope. Sample 13, figure 4, a,consists of fuzzy clumps of short fibers or plates,which is typical of gamma samples. Figure 4, bshows the fine fraction of Cl (delta). Theseparticles are very thin plates. Sample C4 (fig. 5,

4 Sample numbers refer to list of samples in table 3.

a) indicates the fibrous nature of this artificiagamma MnO2 made from the well-crystallizecpyrolusite (sample 12) shown in figure 5, bFigure 6, a and b, shows a cryptomelane beforeand after treatment (samples 11 and C5) showingvery well crysta'lized particles in both casesligure 7, a and b, shovs the extreme aciculaicrystals of cryptomelane formed by autoclaving

Measurement of the specific surface of severalsamples was made by the nitrogen adsorptionmethod [21].5 The samples were evacuated atroom temperature overnight. Nitrogen gas wasadsorbed on the surface at liquid nitrogen tem-perature (—195° C). The surface areas of thesamples were calculated according to the Brunauer,Emmett, and Teller equation [28] and assumingan area of 1G.2 A2 as the area of a nitrogen mole-cule. The results are given in table 4.

TABLE 4. Fineness or M11O2 samples by nitrogen adsorptionmethod

Sample

ClC2C13C7C413

Surface area

myg57.644.140.235.418.57.5

Rating

BA

AAA

As a further measure of particle size, the averagecrystallite size of several samples of gamma anddelta oxides were calculated by the method ofX-ray line broadening [26, 27]. This methodgives the average crystal size rather than the sizeof particles, which may be made up of many smallsingle crystals. The breadths of the peaks on therecorded pattern are measured at one-half inten-sity and compared with the width of similarpeaks of well-crystallized samples (crystals over0.1 /x). This increased width is caused by thesmall crystal size and is used in the calculation asfollows:

0.89XL=where

Be cos 6'

L=average crystallite diameter, A9=Bragg angle\=wavelength of X-ray beam

Be=increased peak width at one-half in-tensity, radians.

5 The specific surface measurement was made by R. F. Blaine of the finenesslaboratory of this Bureau.


at- "H

• • « *

IM.

FIGURE 4. a, Electron micrograph of fine fraction of sample 13 (gamma); b, electron micrograph of sample Cl (delta).

FIGURE 5. a, Electron micrograph of sample C4 (gamma); b, electron micrograph of sample 12 (pyrolusite).


FIGURE 6. Electron micrographs of a sample of cryptomelane.

a, Before treatment (sample 11); b,after treatment (sample C5).

FIGURE 7. Electron micrograph of cryptomelane formed by autoclave treatment.a, Shadowed; b, unshadowed.


The results were as follows:

Sample

1334ClC2C5C3

Averagediameter

A207176117173249220. 5

VI. Heating Curves and High-TemperatureX-Ray Diffraction

Differential heating curves have long been usedto investigate clays and other inorganic solids[22 to 24]. Such methods indicate the tempera-ture, direction, and intensity of thermal changesthat take place when a sample is heated (or cooled)at a constant rate. These changes may be theresult of decomposition, reduction, oxidation, in-version, melting, or other change in phase. Anyparticular compound will undergo certain changes,resulting in peaks on the heating curves that aretypical, just as are the lines of a diffractionpattern. Such curves can then be used empirical-ly as a means of identification and analysis; but ifthe phase change causing the thermal effect can bedetermined, much more can be learned about theproperties and nature of the material. In thisstddy, heating curves of a number of MnO2

samples were made and with some samples X-raypatterns were made at a series of temperatures,making it possible to determine the phase changeinvolved in the heating curve. This was donewith the equipment described above. The pat-terns, in most cases, were made at 100° C intervalsup to about 1,300° C. At each temperature,about 15 min was required to obtain the patternand 20 to 30 min taken between patterns to obtainthe next higher temperature. Thirty-seven heat-ing curves were made; figures 8 and 9 give typicaldata from different samples of MnO2, and table5 shows the phases present at various tempera-tures on certain samples.

On pyrolusite of high purity, such as sample 45(fig. 8, b), the heating curve is quite simple. Thereare endothermic breaks at about 670° C, 950°, and1,170° C. Sample C15 (fig. 8, a) gave similarresults, except that the first break was at a slightly

Modifications of Manganese Dioxide

o

20

60

10020

60 . via 2°_j •ii *

UJ 60Q

10020

60

100100 300 500 700 900 1100 1300 200 400 600 800 1000 1200

TEMPERATURE °C

FIGURE 8. Heating curves of various M11O2 samples.a, No. C15 pyrolusite; b, No. 45 pyrolusite; c, No. 40 pyrolusite; d, No. 3

pyrolusite; e, No. 2 pyrolusite; f, No. 26 pyrolusite; g, No. 13 gamma M11O2;h, No. 46 gamma M11O2.

020

60

100• 20

I 60o ,00

° 6 0

10020

60

100

h

c

d

q

--40-2002040

-40-20

20 oJ40 £-20 S40 u6080

-40-2002040

100 300 500 700 900 1100 1300 200 400 600 800 1000 1200

TEMPERATURE #C

FIGURE 9. Heating curves of various M11O2 samples.

a, No. 42 ramsdellite plus pyrolusite; b, No. 43 cryptomelane; c, No. 44cryptomelane; d, No. 36 cryptomelane; e, No. C14 gamma M11O2; f, No.C2 gamma M11O2; g, No. C8 gamma M11O2; h, No. C13 delta M11O2.

lower temperature. The X-ray data clearly showthat the 600° to 700° C thermal effect is due toloss of oxygen and the formation of bixbyite(Mn2O3), and that the break at about 950° Cis further loss of oxygen and the formation ofhausmannite (Mn3O4).

The thermal change at 1,170° C is more complex.It was known from previous studies that even afterprolonged heating to 1,300° C, diffraction patternsmade at room temperature were of hausmannite,

597

TABLE 5. Results of high-temperature X-ray diffraction studies of manganese dioxide

Tem-pera-ture

°C25

200300400500

600700800900

1,000

1,1001,2001,300

Major phase present for sample—

26

Pyro

Pyro . _

doPyro & Bix.Haus

do,-._-do

doSpinel

do

C15

Pyrodo ..

Bix

Bix

HausSpinel

42

Ram & Pyro. . do

Pyro.

do .Bix

dodo

Haus

doSpinel

46

Gamma-

Pyro

Bixdodo .

Haus.

. . .do—-Spinel

43

Crypt

Crypt

Crypt -Crypt&HausHaus

do.—_..--do -

do

48

Crypt

Cryptdo

. do-

doCrypt & Bix.

doBix

36

Cryptdododo

Crypt--Crypt & HausHaus.--.do—

C2

Gamma .do

Gamma-

Bixdo .dodo

Haus

__.doSpinel

C l l

Delta...

Delta

Bixdodo

Haus

C l

Delta

Delta

Crypt-..-..Crypt&Bix -Bix

doHaus

C8

Gamma.

Crypt.

Crypt&Bix.Do.Do.

Bix.Haus.

Spinel.

therefore no further, reduction was to be expected.It was found that if after reaching 1,200° C, thefurnace was cooled, an exothermic break took placenear 1,100° C with samples of high purity. TheX-ray study indicated that at 1,170° C there wasa rapid and readily reversible inversion from haus-mannite (Mn3O4) to a cubic substance with aunit cell of 8.7 A. This indicates that Mn3O4

(MnO.Mn2O3) forms a spinel structure, isostruc-tural with MgO.Al2O3 above 1,170° C. The spinelstructure is one in which there are many substi-tutions and which occurs under very diverse con-ditions [25]. Hausmannite structure in itself isdefinitely tetragonal but is a distorted spinel struc-ture; thus the inversion at 1,170° C is a simpleone, occurs quickly, and is reversible and thereforeof the a-fi quartz inversion type. The data fromthe pattern made at 1,200° C are given in table 6,along with the room temperature data for bixbyiteand hausmannite.

The heating curve for the sample (42 in fig.9, a) containing the highest percentage of rams-dellite was very similar to that of pyrolusite.There was a section below the first break at 670° Cwhere heat was being absorbed and where theX-ray patterns indicated a change to pyrolusite.This change from ramsdellite to pyrolusite doesnot appear to occur sharply or to be accompaniedby a large energy effect and is not reversible.After the change to pyrolusite, the results werethe same as with pure pyrolusite.

Samples of gamma oxides of natural origin suchas 46 (fig. 8, h) gave heating curves similar to

TABLE 6. X-ray patterns of the lower oxides of manganese

Bixbyite [7;

h k l

200210211220310

222321400

411330420

332422500430

d

A

4.674.203.823.352.99

2.722.512.350

2.206

2.104

2.0041 920

1.869

ri

1010601030

1002040

20

10

4010

40

Hausmannite [7]

h k l

101112200103202

004114

--

-

d

A

4.853.053.872.742.46

2.332.021.809

" 1.775

1.686

1.6261 568

1.534

ri

70502090

100

506010

50

20

1560

80

(No. 42 at l,20(Mn Spine

hkl a

111220311222400

422333440

d

A

5.13.092.632.512.17

1.7691.6681.532

)°C)I

r i

8020

1001515

102018

a Based on cubic with a=8.7 A.

well-crystallized pyrolusite, except that the firstbreak was slightly lower in temperature. Thedata in table 5 indicate a change to pyrolusite,but apparently this was not accompanied by alarge heat effect. After 670° C, the results werethe same as with pure pyrolusite. Sample C8(fig. 9, g) was a very lightweight artificial oxide,which changed from gamma to cryptomelane onheating and then changed over a long range oftemperature to bixbyite. It contained a largeamount of water, which gave the peak near 150° C.

Samples C14 and C2 (fig. 9, e and f), both of

Journal of Research

gamma ore of electrolytic origin, gave patternswith a large exothermic break near 700° C. Thisis thought to result from oxidation of the carbonpresent. But for this effect, the curve would besimilar to other ores.

Cryptomelane and delta samples gave a varietyof types of heating curves. It is understandablethat this would be true when one considers therange of composition of samples with cryptome-lane-type patterns and the fact that these samplesare not pure. The first main temperature effectranges from 550° C in sample 43 (fig. 9, b) to760° C in sample 36 (fig. 9, d), and as a whole thebreak is not as pronounced as with pyrolusite.From 700° to 1,100° C, the curves vary greatly.It becomes evident why this is so when the X-raydata is examined. Some cryptomelane samples,such as 43 and 36, go directly from cryptomelaneto hausmannite with no intermediate stage ofbixbyite. Others change gradually to bixbyite (asCl and 48). This is probably due to the effect ofcertain impurities on the transformation. In onecase (sample 26), a sample of pyrolusite gave adouble break (fig. 8, f) between 600° and 700° Cand showed formation of hausmannite with onlya partial fomation of bixbyite. Some of the extrathermal effects at high temperatures were unex-plained. Sample 43, for example, gave an un-identified phase above 1,200° C. Fusion causessome of the extra breaks at high temperature incases of certain samples.

VII. Conclusions

1. There are five major types of manganesedioxide of importance to battery technology: (a)Pyrolusite; (b) gamma MnO2, a form of pyrolusiteof fine crystal size, (c) ramsdellite, (d) crypto-melane, a compound of variable composition withessential K or Na, (e) delta MnO2, apparently afinely crystalline form of cryptomelane.

2. The heating curves of the various types fallinto two general groups, one containing pyrolusite,ramsdellite and gamma, the other, cryptomelaneand delta.

3. The latter group gives a greater variety ofheating curves, partly because of a greater rangeof composition.

4. Impurities, which may or may not affect theoxides for battery use, have a strong influence onthe heating curves, which makes it doubtful if this

test can be of definite value, used alone, to evaluateoxides.

5. Gamma samples break down to bixbyite ata lower temperature than well-crystallized pyro-lusite.

6. The shape of the fine fractions as seen in theelectron microscope is distinctive.

7. The specific surface of commercial samplesvaries widely, being highest on artificial oxides.

The authors thank the U. S. Geological Surveyand the commercial battery manufacturers forfurnishing samples of manganese dioxide. Theyare also grateful to E. S. Newman of this Bureaufor help in obtaining the heating curves, and toG. W. Vinal for many helpful suggestions.

VIII. References

[1] H. F. McMurdie, Microscopic and diffraction studieson dry cells and other raw materials, Trans.Electrochem. Soc. 86, 313 (1944).

[2] W. F. Cole, A. D. Wadsley, and A. Walkly, An X-raydiffractioxi study of manganese dioxide, Trans.Electrochem. Soc. Preprint 12-2.

[3] A. Van Valkenburg, Jr. and H. F. McMurdie, Hightemperature X-ray diffraction apparatus, J. Re-

' search NBS 38, 415 (1947) RP1782.[4] S. Speil, L. H. Berkelhamer, J. A. Pask, and B. Davies,

Differential thermal analysis—its application toclays and other aluminous minerals, U. S. Bur.Mines, Tech. Pap. 664.

[5] E. S. Newman and L. S. Wells, Effect of some addedmaterial on dicalcium silicate, J. Research NBS36, 137 (1946) RP1696.

[6] A. Ferrari, L'esame rongtenographico, dei reticolicristallini del fluoruro manganoso e del biossido dimanganese, Acad. naz. Lincei 3, 224 (1926).

[7] M. Fleischer and W. E. Richmond, The oxides ofmanganese, Econ. Geo. 38, 269 (1943).

[8] O. Glemser, Uber eine neue Modification des Mangandioxyds, Ber deut. chem. Ges. [B], 72, 1879 (1939).

[9] F. Schossberger, Uber die rontgenograph Untersuch-ung von naturlichen un kunstlichen Braunstein,Physik. Ber. 22, 1340 (1941).

[10] A. St. John, The crystal structure of manganese di-oxide, Phys. Rev. 21, 389 (1923).

[11] J. W. Gruner, Groutite, HMnC>2, a raw mineral of thediasporegoethite group, Am. Mineral 32, 654 (1947).

[12] J. W. Gruner, The chemical relationship of crypto-melane (psilomelane) hollandite and coronadite,Am. Mineral 28, 497 (1943).

[13] C. Frondel and E. W. Heinrich, New data on hetaero-lite hydrohetaerolite, coronadite and hollanditeAm. Mineral 27, 48 (1942).

[14] W. E. Richmond and M. Fleischer, Cryptomelane, a


new name for the commonest of the "Psilomelane"minerals, Am. Mineral 21, 607 (1642).

[15] L. S. Ramsdell, The unit cell of cryptomelane, Am.Mineral 27, 611 (1942).

[16] C. Palache, H. Berman, and C. Frondel, The systemof mineralogy, 7th ed. (John Wiley and Sons, Inc.,New York, N. Y., 1935).

[17] W. Feitknecht and W. Marti, Uber die Oxydation,von Mangen hydroxyd mit molekalaren Sauer-stoff, Hel. Chem. Acta 28, 129 (1945).

[18] M. P. DuBois, Contribution a l'etude des oxyden duManganese, Ann. de Chimie 5, 411 (1936).

[19] Von Julius Meyer and R. Kanters, Zur Kenntnis desdreiwertagen Mangans VII, Z. anorg. Chem. 185,178 (1929).

[20] H. B. Weiser, Inorganic colloid chemistry, II, thehydrous oxides and hydroxides (John Wiley andSons, Inc., New York, N. Y., 1935).

[21] P. H. Emmett, Symposium on new methods forparticle size determination in the subuene range,Am. Soc. Testing Materials (March 1941).

[22] S. Cailiere, S. Henin, and L. Tane, Investigations of

the differential thermal analysis of clays, Compt.rend. 323, 383 (1946).

[23] R. E. Grim and R. A. Rowland, Differential thermalanalyses of clay minerals and other hydrous ma-terials, Am. Mineral 27, 746, 801 (1942).

[24] R. E. Grim, Differential thermal curves of preparedmixtures of clay minerals, Am. Mineral 32, 493(1947).

[25] R. W. G. Wyckoff, The structure of crystals, 2d ed.,p. 290 (Chemical Catalog Co., New York, N. Y.,1931).

[26] M. H. Jellinek and I. Fankuchen, X-ray diffractionexamination of gamma alumina, Ind. Eng. Chem.37, 158 (1945).

[27] G. H. Cameron and A. L. Patterson, Symposium onradiography and X-ray diffraction, Am. Soc.Testing Materials, p. 324 (1936).

[28] S. Brunauer, P. H. Emmett, and E. Teller, Absorp-tion of gases in multimolecular layers, J. Am.Chem. Soc. 60, 309 (1938).

WASHINGTON, June 11, 1948.


Documents

By Howard F. McMurdie and Esther Golovato I. Introduction