U. S. Department of Commerce Research Paper RP1941 National Bureau of Standards Volume 41, December 1948 Part of the Journal of Research of the National Bureau of Standards
Study of the Modifications of Mangan,ese Dioxide By Howard F. McMurdie and Esther Golovato
P ast work on the modificat ions of manganese dioxide of in terest in dry-cell manufacture is revie wed. New X -ray data, at both room and elevaLcd temperatures, combined 'with differen tial hea ting curves lead to the concl usion that five type of manga nese dioxide exi t : (1) well-crystallized p yrolusite ; (2) gamma m a nganese diox ide, a poorly crystalli zed pyrol u site; (3) ramsdellite ; (4) cryp tomelane, a form co ntaining esse ntial p otassium or sodiulll.; a nd (5) delta m anganese dioxide, belie ved to be a poorly crystallize d cryptomela ne. The high-temperature X-r ay diffraction data indicated the phase cha nges t hat cause the heating
curve effects. A new crystal fo rm of manganosic oxide (Mn30 4), stable above 1,1700 C, w ~ s fo und t o be cubic of spinel structure. Fincness dcten n inaLion by both thc nitrogen a dsorp t ion and Lhe X-ray line broadening me thods were made on selected samples.
I. Introduction this equipment a fla t specinlen is used, and no During the years 1940- 46 there wa increased special teclmiq Ll es were employed to prcven t pre researeh on dry cells. This was stimulated by fcn 'cd orientation. It is r ealized that in a few increased demand for the cell as well a new u es cascs this may have res ulted in r elative intcnsitics for them , combined with certain hor Lages of raw that differ from those in other r eporLs. This materials. This work disclosed among other equipmen t in it commercial form is not capable t-hings that mancr ane e dioxide i not a simple of recording the diffracLion eff ects at angles grcater compound with consLant properties, and that it than 45 ° 8; thus, the bacl rcfl cc Lion lines are value as a depolarizer depends on pl'oper ties other mi ed. ome conclusions arc bascd in par t on than merely purity. At that time a paper was previous Ludies with thc use of photographic written at this Bureau [1] 1 criving some prelim m ethods. inary findings. Since then, addiLional work has Thc X-ray diffraction p atterns at elevated tem b een done both h ere and elsewhere [2] on man perature were made in the appara Lu describcd ganese dioxide. The present paper is an attempt by Van Valkenburg and M cMurdie [3]. With t o evaluate the work done here and that reported this, pattcrns could be obtained at variou ele by others. The work h a not been confined to vated tcmperaturc with no intermediate cooling. dry-eell technology, but was aimed toward a In many measurcments at high temperature, only better general understanding of the oxide, its the section of the pat tcrn corresponding to angle v arious forms, transformations, and means of less than 25° 8 was scanned, as this was found to iden tification. This information, it is hoped, will be sufficient to iden tify thc phase present. be of interes t in mineralogy and crystallography Differential heating curves on Mn02 were made as well as in electrochemistry. on 2-g samples by the m ethod outlined by p eil and others [4] . Al20 3 (corundum) was used as an II. Apparatus and Methods inert body. The temperature of the Mn02, and In the present study, X-ray pa tterns were made the difference in temperature b etween the Mn02 on the North American Philips Geiger Counter and the inert body were automatically recorded X -ray Spectromet er by using unfiltered F eK radi as the tempcrature was raised at a rate of about ation. The patterns were automatically rccorded 8° C/min. The equipment used was de cribcd by with a counter movemcn t of 1 ° 8/min. With N ewman and Wells [5]. Electromnicrographs were made with th e RCA I F igures in brackets indicate the lite rature references at the end of t bis paper . model EMU microscope.
.Modifications of Manganese Dioxide 589 III. The Crystal Modifications of Mn02 Although there is more general agreement on the - forms of Mn02 now than existed several years ago, a there still is a certain amount of disagreement on II the relations between the various forms. For ex ample, all agree that gamma Mn02 is a poorly crystallized material, but Cole and his coworkers > .... b [2] claim that it is closely related to ramsdellite, if) z I ,I I , whereas McMurdie [1] has considered it related to W f pyrolusite. As a first step toward settling these Z
questions, it must be agreed as to what lines in w the X -ray diffraction pattern are essential to de > ~ I , lineate a particular modification. -l .W There is general agreement on the well-crystal 0:: lized tctragonal form of Mn02, pyrolusite. This was made artificially by Ferrari [6] and the crystal structure worked out. It was found to be of rutile I. , I. structure with a= 4.44 A and c= 2.89 A. The lines for angles less than 40° 0, which are present in an X-ray pattern, using Fe radiation, are given in table 1. It will be seen th at some of the lines , , , I reported by Fleischer and Richmond [7 j are 024 6 8 10121416182022242628303234363840 I DEGREES 9
FIGURE 2. X-my diffractwn pattem .~ of various samples of 1\1]10 2, I a, ~TO. C15, p yrolusite; b, Ko. C9 gamma JVfn02 plus cryptomelane; c, No. I. .1 011 delta MnO" d, No. 46 gamma MoO,; e, No. 08 gamma MoO,.
mlssmg. These extra lines arc not compatible with the structure given by Ferrari and arc not > found here in the purest samples. They are .... I \ ,1 , I ~ ~, ~--~~~~~~~~~~~~~~~~~~~I. I doubtless due to impurities. In figure 1, a/ the ....lJJ pattern of pyrolusite obtained from a sample Z of natural ore from Egypt is represented. The llO and the 220 lines are possibly unduly strong because of prefcrred orientation in the flat holder. Figure 2, a, sh ows the pattern obtaincd from arti ficial material made by heating Mn(N03)2' This pattern is similar to that of the one in figure 1, a, I, II II except for the increased broadness of the lines due / to smalle[' crystal size. The mineral name (( polianite" was used for many years to indicate the well-crystallized form, as distinct from massive material called "pyrolu site". It has been shown beyond question [10] by o 2> 4 6 llO 1'2 14 I~ 18 2'0 2'2 ~4 2~ 28 3032 3'43'6 38 40 DEGREES 9 X-ray diffraction that the two are identical and the name polianite js now discarded. It was a FIGURE 1. X -my di.ffmction pattems of various modifica tions of 2• 1\1n O 2 The 0 (theta) values indicated on these figures arc the angles of din'raction a, No. 45 pyrolusite; b, N o. 42 ramsdcllite plus pyrolusite; c, No. 016 gamma and are related to the d values shown in the tables by the Bragg equation Mn02; d , No. 40 cryptomelan e; c, No. Cl delta :.\1n02. d=),, /2 sin 0, where ),,=wfivelength of F eK a·radiation (l.936 A).
590 Journal of Research TA TILl;; 1. X-ray powder data use. The patterns reported for it vary, som e workers cataloging several varieties [2]. The Pyrolusite (No. 45) Oanuna (No. CIG) Ramsclel1ite f2] pattern reported by McMurdie [1] has been eriti
hkl d In hkl II Rl" d Rl eized as not including a line at a spacing of about ------4 A (about 14°0). Although it is true that in A A A many cases, such as shown in figure 2, b, and figure 110 3.14 100 110 3.2 GO 4.04 Very strong. 101 2.42 80 10l 2.4 100 2.54 Strong. 2, d, such a line does occur with a pattern that is 200 2.22 10 2.43 Medium. ?t~erwise mainly that of gamma, in other cases III 2.12 50 III 2.lt 70 2.33 Do. 210 1. 985 10 2.13 Medium strong. It IS much weaker or absent. This would seem 21 1 1. G35 80 211 1. 63 80 2.05 Very weak. to indicate that it is not caused by the same phase 220 1. 570 30 1. 90 Medium. 1. 65 Medium strong. as produces the gamma pattern and is not es 1. 6t Do. sential to it. The pattern shown in fio'ure 1 c is . b " of an OXIde made by pouring concentrated -All lines broad. Mn(N0 3)z solution drop wise onto a hot plate near 400° C. This, except for th e lack of the 4-A line common boli ef years ago that the mineral that is approximately what is accepted as a 'was without external crystal faces (then called gamm ~ pattern. When Lhis is compared with the pyro pyrolusite) was superi or for dry cells, because it lu slLe paLtern, it is seen LhaL it is similar and 'was hydrated. There is at present no evidence of varie in having broader lines (ind icating smaller a hydrated form of Mn02, no other essen tial crysLal size) and the absence of the 200, 2J.0 and clements being present. 220 lines. This would indicate a subs Lance 'very l;Yad is a name that h as been used to designate poorly developed perpendieular Lo the c axis, and a massive or earthy mixture that is largely Mn02, of very small crystal size. The cause of Lhe li ne but it often contains large percentages of H 20 and at 4 A, wl1 i h occurs in some cases is not known ' BaO. It has no definite mineralogical meanincr . ' - , It may be related Lo ramsdellite 01' grouLite and actually consists of pyrolusite, psilomelan:' (HMn02) repor Led by Gl'uner [11]. Bo Lh of th ese cryptomelane, and other minerals [16]. phases give sLrong lines aL a bOll t 4 A and may be A second well-crystalli zed form, which was ~r esenL 1Il pood? crystallized form with Lhe poorly reported first by Fleisch er and Richmond [7] i formed pyroluslte (gamma). The fact Lhat vari ramsdellite. This is said to be or Lhorhombic, but ous differences in th e gamma 1-'In02 paLLern occur the unit cell dimensions have not been given. is not surprising, as the degree of crystallization Cole and coworkel's [2] reported a paLLern for Lhis a~d p~rc enLage of a second phase may vary over material that can be derived from that of Fleischer WId e lImIts, thus giving a considerable range of and Richmond by subtracting the lines of pyro patterns. lusite from the latter. Of the various samples It is though t very unlikely that the differences of ramsdellite studied here, all have eontained sueh as noted by Cole and eoworkers [2], are caused pyrolusite in varying degrees. It is believed br.real difIerences in stru cture; they are probably therefore, that the best pattern of ramsdellite i ~ dIfferent In degree of crystallization and of im that reported by Cole. This is reprin ted in table purities. The pattern, as shown in fiO'Llr e 1 e is 1. The paltern of the purest ramsdellite obtained b e1 ·LOve d to show the minimLlm numbeb"r of lines here is shown graphieally in figure 1, b. Rams that delineate this modification of 11nOz. dellite is llOt known to have been made artificially. Cryptomclane occurs as a natural mineral, has The modifieation of the greatest interest in been prepared (synthetically), and found to be battery technology, and about which the most isostrucLural with hollandite and coronadi te [12 confusion has arisen is gamma (-y) 3 MnOz. This 13]. It is th e alpha MnOz of Du Bois [18]. The s~ was first named by Glemser [8] and has been minerals have been given the gen eral formula noted by many workers [9, 1]. It is found in R l\fn 01 6' R is K or N a in cryp tomclane Pb in natural ores and in artificial produeLs and is s coronadite, and Ba in hollandite. The' name generally considered a desirable form for battery cryptomelane was proposed for the alkali-bearino' 3 The terms gamma and della used here in connection with t,-pes of MnO co ~pound by Richmond and Fleischer [14], i~ are ~ot c~nsldered to refer to distinct mineral form s, but rather to denote modificatIons of technical interest. bemg th e commonest of the minerals of tIllS type.
Modifications of Manganese Dioxide 591 The unit cell as determined by Ramsdell [1 5] was found to be body-centered tetragonal with a= 9.82 0 A and c= 2.86 A. The pat tern given in table 2 I I. and shown in figure I , d, agrees with this cell.
T A BLE 2. X-my powder data
~ I, I , ,II ,II Cry ptomelane (No. 43) Cryptomelane (:-;ro . 43) ':::: (j) z w hkl d RI h kl d RI >------~ A w A ;::: 110 6. 9~ 90 210 2. 2t 20 I, 4.9 l W :101 2. 16 25 ~ 200 ...J 3.47 30 111 I. 8:l5 20 W " 220 J 30 a: I ~O :l. ll IGO 600 1.6'1 400 2 ..jr. 20 251 1. 54 15 121 2. 40 4U I, I, P silomelanc is a name that has been used in connection with manganese dioxide minerals, and various formulae have been proposed for it. The most recent usage restricts psilomelane to a , I, ,ii , I , , , ,I ,I , hydrous barium compound (BaMnMns0 16 (OH)4) o 246 8 101214 16182022242628303234363840 [16] identical with romanechite. This mineral DEGREE S e does not appear to be common among oxides F IGU R E 3. X -ray diffraction patterns of various samples of proposed for battery use. Its X-ray pattern, Mn02. however , is similar to that of cryptomelane and a, No. 34 gamma M n O, p lus cryptomelane; b, :-;roo C5 cryptomelane; c, No. C9 gamma M nO, plus crYl)tomelane; d, No. 39 delta M nO,; e, No. 26 small amounts of psilomelane in mixtures are pyrolusite. easily mistaken for cryptomelan e. D elta Mn02 is a modification, similar to gamma bears to cryptomelane a similar relation as that of Mn02 in that it is poorly crystallized and that gamma to pyrolusite in being very finely crystal th ere is conflicting data on th e essen tial lines. line; and since various degrees of crystallinity can This modificat ion was so designated by M cMurdie exist, the pattern will vary. [1] and is believed to be th e sam e as the material Consideration has been given to the possibility called amorphous by Gruner [12] and manganous of a hydrated form of Mn02. Such a compound m anganite by F eitnecht and Marti [17] . Cole and has not been found. Weiser [20] made Mn02 coworkers [2] prepared a sample giving the two samples by various wet m ethods, and although no lines reported by M cMurdie (at 2.41 and 1.42 A) X-ray data are given, from samples prepared by and also a type with more lines, particularly one similar m ethods by others, it is known that t he l at 7.13 A. These samples in many cases could be samples must have included delta and gamma i changed to Cl'yptomelane by boiling or heating. specimens. These were dried at various tempera- ~ With the Geiger-counter equipment in use here at tures, and he states that no evidence was found the moment, t h e patterns of various samples have that hydrates exist and that the water of the finely been repeated. This apparatus does not indicate divided forms is adsorbed. th e lines with as sm all interplanar spacings as was IV. Occurrence of the Modifications observed by the film m ethod formerly used, but does give bet ter delineation of lines at greater Patterns l1ave been m ade here on samples, both spacings. The fewest lines found on a sample in artificial and natural, that hav.e been proposed for th e () range covered are shown in figure I , e. or used for battery depolarizers. The principal Another pattern is shown of a similar sample with Mn02 modification, as determined by X-ray dif additional lines in figure 2, c. One sample of fraction, and the source of the sample are given delta Mn02 was found occurring naturally and is in t able 3. The m ethod of manufacture and of shown in figure 3, d . The evidence is that delta trea tm ent of some commercial orcs is not known.
592 Io.urnal of Research ~----__~r------~
IAlthough it would add to the value of this report TABLE 3. Samples of 1Iwnganese dioxide- Conlin ued [if these da ta w('re known , it would also be of value
to have exact comparative data on the behavior of P attern Perc('lIt Sample Source ageo f R Rtin g these ores in cell . In many cases, however, data MnO, ~;at u ra l samples are available by which ore can be roughly clas ------1------1------Isified. In column 5 of table 3, ratings are given on 41. ______Lower Oalifornia __ Pyro+crypL _ .__ 76 0 42 ______New Mexico ______R
Artificial samples T A BLE 3. Samples of manganese dioxide
01. ______OhemicaL ______D elia ______B 02 ______E lectrolytic ______Gamma ______A Pattern Pereent- Sample Source 1------1 a~e of Rating· 03 ______No. 10 "treated" ______do ______MoO, A Natu ral samples C4 ______No. 12 Ht rea ted" ______do ____ . ______A ------1------1--- 5 ______No. 11 " treated" __ OrypL ______B C6 ______No. 13 "treated" __ Gmnma ______1. ______New Brunswick __ Pyro ______72 o A 07 ______Electrolytic __ . ____ Gamma+crypt ______2______Java ______do ______86 o A 3 ______. Mexi co ______. ______do ______81 D 08 ______4 ____ . ____ Nova Scotia ______do. ______OhemicaL ______Gamma ______83 o 09 ______do ______Galllma+crypt ______5______Utah ___ . ___ . ______. do ____ . -- ___ _ 82 D 011. ______do ______DelL,L_. ______A 013 ______do ______do ______6 ______Arizona ______. __ do ______78 o 014 ______Electrolyti c ______Gamma ______7______Mexico ______p yro+crypL __ _ 78 D A 8 ______Australia_. ___ .. ____ OrypL ______81 .B OJ5 ______P yro from Pyro ______9 ______Oaliforcia ______Gamma ______68 A 10 ______Moroceo ______Pyro ______91 D Mn(_ 0 3),. 016______Gamma from Gamma. ______. ______ll ______Mexico ______Orypt . ______78 o Mn(N0 3),. 12. ______Cuba ______Pyro ______o 13______Gold Ooast______Gamma______Ii" A 14 ______Greece . ______do______62 n 15 ______._ Arkansas ______Pyro______82 o In this work no attempt has been made to prepare material in various ways, but from the 16 ______Mexico ______do______61 D 17 ______Phillipines __ . ______do______69 o findings reported here and in previous papers cer 18 ______Virginia ______CrypL______73 B tain conclu ions can be draw'n. 19 ______Java ____ . ______Pyro ______88 B 20 ______Papua ______Gamma __ .______88 A Pyrolusite (well crystallized) occurs commonly in nature and can be prepared ar tificially by 21. ______Soutb America ____ Pyro______78 o 22 ______Ouba ______do ______D beating Mn (N0 3)2; by hydrolysis of MnCl4 [2], 23 ______Morocco ______do______87 o and by heating cer tain cryptomelane or delta 24 ______Montana ______do______83 o 25 ______Afriea ______Gamma+crypt _ 74 A samples [2]. Gamma samples become better crystallized and form pyrolusite on heating [2 , ]. 26 ______Oaueasia ______Pyro______86 B ·27 ______Oalifornia ______Gamma______66 A It has been shown by several investigators [2 , 7] 28 ______lndia ______Pyro+crypL___ 84 o that ramsdellite changes to pyrolusite on heating 29 ______M ontana ______Orypt ______. ------B 0 30 ______BraziL ______Pyro ______------near 500 C. Gamma manganese dioxide (poorly crysta.llized 3L ______Nel' ada ______do______67 o 32 ______Wyoming ______do______83 o pyrolusite) occurs to some oxtent in nature. The 33 ______Obina ______. ____ do______81 o well-known Gold Coast ore is largely of this form, 34 ______Runan Obina ___ __ Gamma+erypt ------A 35 ______I ndia ______Pyro+erypL___ 83 o and other occurrences are listed in table 3. The gamma manganese dioxide was prepared here by 36.______Utah______OrypL______69 n heating a saturated solution of Mn (J0 )2 37 ______Nova Scotia ______P yro______88 o 3 38 ______do ______do______88 B quickly and is of common occurrence in electro 39 ______Oanada ______Delta______68 B 40 ______Morocco ______Pyro______91 D lytically prepared samples. Cole [2] prepared
I Modifications of Manganese Dioxide 593 gamma cllemically by various means but was a) indicates the fibrous nature of this artificial unable to foretell when gamma and when cryp gamma Mn02 made from the well-crystallizec; tomelane would result. The patterns of samples pyrolusite (sample 12) shown in figure 5, bl C3 and C4, table 3, indicate that the "treatment" Figure 6, a and b, shows a cryptomelane befon used formed gamma from well-crystallized pyro and after treatment (samples 11 and C5) showinE lusite. The treatment was done by a commercial very well crystallized particles in both cases! battery company and is not known to us. From ] igure 7, a and b, sho !Vs the extreme aciculaI discharge data available, it is evident that the crystals of cryptomelane formed by autoclavingl reduction of crystal size (and thus increased :Measurement of the specific surface of several surface) resulting from the process improved the samples vvas made by the nitrogen adsorptior ore for battery use. method [21].5 The samples were evacuated a~ Cryptomelane has been made by Cole [2] and in room temperature overnight. Kitrogen gas wa2 this laboratory by precipitation of Mn02 from adsorbed on the surface at liquid nitrogen tem . KMn04 solu tion. In some cases, the poorly perature (-195° C) . The surface areas of thJ crystallized delta Mn02 resul ted, and cryptomelane samples were calculated according to the Brunauer, was made from this by heating. Cole [2] found Emmett, and Teller equation [28] and assumin~ that electrolysis of solutions containing K or Pb an area of IG.2 A2 as the area of a nitrogen mole gave a matcrial with a cryptomelane pattern. cule. The results are given in table 4. . (The pattcrn of coronaditc is similar to that of cryptomelane and may be the form resulting in TABLE 4. Fineness or 1\1n02 samples by nitrogen adsorptiol method the case of Pb). Sample C5,4 with presumably
the same treatment as samples C3 and C4, gave a Sample Surface area R ating cryptomelan e pattern, just as did the untreated material. Cryptomelane, of a,n extremely fibrous m'/(I 0 1 ...... •••...... •. 57. G B nature, was found to result from a.utoclaving 02. ______. .... ______.______44.1 A 013 ______.. ______. 40.2 certain samples of delta and gamma Yfn02 [1] . 07. ______3.5.4 A In nature cryptomelane is very abundant [7]. 04. ______. 18. 5 A 13 ______. 7.5 A Delta Mn02, believed to be poorly crystallized cryptomelane is found rarely in nature, only one L....~~~~~~~~~'___~~_'____~~_I I such sample being available here (sample 39, fig. As a further measure of particle size, the average 3, d) , but it is formed from KMn04 solution by crystallite size of several samples of gamma andl precipitation with HCI. It is found as a major delta oxides were calculated by the method of, constituent in many chemically produced oxides. X-ray line broadening [26, 27]. This method Ramsdellite is known to occur with certainty gives the average crystal size rather than the sizel only at Lake Valley, N. Mex. Here it is inti of particles, which may be made up of many small mately related to the pyrolusite. It is not known single crystals. The breadths of the peaks on the to have been made artificially. recorded pattern are measured at one-half inten-' sity and compared with the width of similar) V. Particle Size and Shape peaks of well-crystallized samples (crystals over' Because the action of Mn02 in dry cells is 0.1 J.l ). This increased width is caused by thel largely a surface action, the degree of fineness is small crystal size and is used in the calculation aS I important. The gamma and delta Mn02 are follows: finely divided, as shown by the line broadening in L 0.89A the diffraction patterns. This fine crystal size Be cos e' where and high surface area arc also shown by the L = average crystallite diameter, A electron microscope. Sample 13, figure 4, a, 8 = Bragg angle consists of fuzzy elumps of short fibers or plates, A= wavclongth of X-ray beam which is typical of gamma samples. Figure 4, b Be= increased peak width at ono-half in- i shows the fine fraction of C1 (delta). These tonsi ty, radians. particles arc very thin plates. Sample C4 (fig. 5, 5 'I'hc specific surface mensllremeot was made by R. F . Blaine of the fin eness • Sample num bel'S refer to list of samples in table 3. laboratory of this Bureau.
594 Journal of Research
I I "'-- FIGUlm 4. a, Electron microgmph of fine j?-action of sample 13 (gamma); b, electron micrograph of sample C1 (delta).
FIGUHE 5. a, Electron micrograph of sample C4 (gamma); b, electron micrograph of sample 12 (pyrol1lsite).
Modifications of Manganese Dioxide 595 FIGURE 6. Electron micrographs of a sample of cryptomelane. a, Before treatment (sample 11); b , after treatment (sa mple C5).
F IGURE 7. Electron microgmph of cryptomelane formed by autoclave treatment. a, Shadowed; b, unshadowed.
596 Journal of Research I The results were as follows:
Average Sample diameter
A 60 13 207 z 34 176 o C1 117 ~ 100 r-----~~------~~------~ C2 173 ~ 20 ('5 249 ...J . C3 220. 5 ~ 60 o
VI. Heating Curves and High-Temperature 6 0 X-Ray Diffraction 100 ~~~~~~~~~~~~~~~~~~~~ Differential heating curves have long been used 100 300500 700900 11 00 1300 200 4{)0 600 800 1000 1200 to investigate elays and other inorganic solids TEMPERATURE ·C [22 to 24]. Such methods indicate the tempcra FICU RE 8. !Tealing curves oJ vari01ls iVln02 samples . ture, direction, and intensity of thermal changes a, No. 0 15 p yro lusite; b , N o. 15 pyrolusite; c, No. 40 pyrolusite; d, No.3 that take place when a sample is heated (or cooled) p yrolusite; e, No.2 pyrolusite; f, No. 26 pyrolusite; g, N o. 13 gamma MuO,; h , I 0. 46 gamma MoO,. at a constant rate. These changes may be the
result of decomposition , reduction, oxidation, in 0 -40 20 version, melting, OT other change in phase. Any -20 0 particular compound will undergo cer tain changes, GO 20 4 0 resulting in peaks on the heating curves that are 100 -40 typical, just as arc the lines of a difrraction 20 -20 0 pattern. Such curves can then be used empirical z 60 z 0 20 0 ly as a means of identification and analysis; but if >= 40 >= " 100 the phase change causing thc thermal effect can be '" -20 "..J ~ 20 '"U. 0 40 '" determined, much more can be learned about the '" 60 60 0 properties and nature of the material. In this 80 10 0 - 40 stuely, heating curves of a number of Mn02 20 -20 samples were made and with some samples X-ray 0 60 patterns were made at a series of temperatures, 20 making it possible to determine tbe phase change 100 300 500 700 900 11 00 1300 200400 600 800 1000 1200 involved in the heating curve. This was done TEMPERATURE 'C with the equi.pment described above. The pat FICURE 9. JJe at~ng curves oJ val'ious iVln02 samples. terns, in most cases, were made at 100° C intervals a, N o. 42 ramsdcllite plus pyrolnsite; b, N o. 4.3 eryptomela ne; e, No. 44 up to about 1,300° C. At each temperature, cryptomelaue; d, N o. 36 cryptomelane; e, N o. 0 14 gamma MnO,; f, No. about 15 min was r equired to obtain the pattern 02 ga mma MnO,; g, N o. 08 gamma MnO,; h, N o. 0 13 delta MnO,. and 20 to 30 min taken between patterns to obtain the next higher temperature. Thirty-seven heat lower temperature. The X-ray data clearly show ing curves were made; figures 8 and 9 give typical that the 600 0 to 700 0 C thermal effect is due to data from different samples of Mn02, and table loss of oxygen and the formation of bixbyite 5 shows the phases present at variolls tempera (Mn20 3) , and that the break at about 950 0 C tures on certain samples. i further loss of oxygen and the formation of On pyrolusite of high purity, such as sample 45 hausmannite (Mn30 4)' (fig. 8, b), the heating curve is quite simple. There The thermal change at 1,170° C is more complex. are endothermic breaks at about 670° C, 950°, and It was known from previous studies that even after 1,170° C. Sample C15 (fig. 8, a) gave similar prolonged heating to 1,3 00 ° C, diffraction patterns results, except that the first break was at a slightly made at room temperature were of hausmannite,
Modifications of Manganese Dioxide 597 TABLE 5. R esults of high-temperat1lre X-ray diffracl,ion studies of manganese dioxide
T em· Major phase presen t for sample- pora· 1---- -;----,-----.,..--...... ,------,------,--- --...,-----;---,------,----1 turc 26 C15 I 42 I 46 43 48 36 C2 C ll C1 C8 --- :----: °C 25 P yro ... __ __ Pyro____ Ram & Pyro Gamma. Crypt. ______Crypt______Crypt.______Gamma_ D elta__ _ Delta______Gamma. 200 ______d o ______do _. ______do ______do ______300 ______Crypt. ______do ______Delta ______400 ______Crypt ______do ______do ______Gamma ______Delta ______500 P yro ______Pyro ______Pyl'O ______do ______Crypt.
600 _____ do ______do ______Cryp!.. ______do ______BiL __ __ CrypL _____ Crypt&Bix. 700 P yro&BiL BiL ____ BiL ______BiL ____ Crypt&Haus Crypt & Bix. CrypL ______do _ ___ Bix. __ __ Crypt&Bix _ D o. BOO H a Ll s ______do ______do _____ H aus ______do ______Cryp t&EaGs ___ do ______do _____ BiL ______Do. 900 __ __ _do ______BiL ______do ______do ______do ______BiL ______Raus ______do ______do. ______do ______Bix. 1,000 __ ___ do ______H aLls ______Haus. ______do ______do ______Haus _. ______Haus ______Haus.
J, 100 __ ___ do ______R aus ______do ______do ______do ______do ____ _ Hans ______J, 200 SpineL _____ Spi neL_ SpineL ______SpineL ______SpineL ______Spinel. 1. 300 ___ __ do ______. ______
therefore no further. reduction was to be expected. TABLE 6. X-ray patterns of the lower oxides of manganese It was found that if after reaching 1,200° C, the (No. 42 at 1.200 0 C) furnace was cooled, an exothermic break took place Bixhyite [7] I H ausm annite [7] Mn Spinel near 1,100° C with samples of high purity. The X-ray study indicated that at 1,170° C there was hk~ l d I ri ~ l __d __ ri_ ~~ _ _d _ _ r_i _ a rapid and readily reversible inversion from haus A A A mannite (Mn30 4) to a cubic substance with a 200 4.67 10 10J 4.85 70 111 5. 1 80 unit cell of 8.7 A. This indicates that Mn30 4 210 4.20 10 112 3.05 50 220 3.09 20 211 3.82 60 200 3.87 20 311 2.63 JOO (MnO.Mn203) forms a spinel structure, isostruc 220 3.35 JO 103 2.74 90 222 2.51 J5 tural with MgO.Al20 3 above 1,170° C. The spinel 31 0 2.99 30 202 2. 46 100 400 2.17 J5 structure is one in which there are many substi 222 2.72 100 004 2. 33 50 422 1. 769 10 tutions and which occurs under very diverse con 321 2.5J 20 114 2.02 60 333 1. 668 20 ditions [25]. Hausmannite structure in itself is 400 2. 300 40 -- 1. 809 10 440 1. 532 18 411 2.206 20 1. 775 50 ------definitely tetragonal bu t is a distorted spinel struc 330 -- --- ture; thus the inversion at 1,170° C is a simple 420 2.104 10 -- J. 686 20 ------
one, occurs quickly, and is reversible and therefore 332 2.004 40 - - 1. 626 15 ------of the a - {3 quartz in version type. The data from 422 I. 920 10 - - 1. 568 60 ------500 1. 869 40 1. 534 80 ------the pattern made at 1,200° C are given in table 6, 430 -- - - - along with the room temperature data for bixbyite and hausmannite. • Based on cubic with a=8.7 A. The heating curve for the sample (42 in fig. 9, a) containing the highest percentage of rams well-crystallized pyrolusite, except that the first dellite was very similar to that of pyrolusite. break was slightly lower in temperature. The There was a section below the first break at 670° C data in table 5 indicate a change to pyrolusite, where heat was being absorbed and where the but apparently this was not accompanied by a X-ray patterns indicated a change to pyrolusite. large heat effect. After 670° C, the results were This change from ramsdellite to pyrolusite does the same as with pure pyrolusite. Sample C8 not appear to occur sharply or to be accompanied (fig. 9, g) was a very lightweight artificial oxide, by a large energy effect and is not reversible. which changed from gamma to cryptomelane on After the change to pyrolusite, the results were heating and then changed over a long range of the same as with pure pyrolusite. temperature to bixbyite. It contained a large Samples of gamma oxides of natural origin such amount, of water, which gave the peak neal' 150° C. as 46 (fig. 8, h) gave heating curves similar to Samples 014 and C2 (fig. 9, e and f), both of
598 Journal of Research L_ gamma ore of electrolytic onglll, gave patterns test can b e of definite value, used alone, to evaluate with a large exothermic break near 700 0 C. This oxides. is though t to fe ult from oxidation of the carbon 5. Gamma samples break down to bixbyite at present. BuL for th is eff ect, the curve would be a lower temperature than well-crystallized pyro similar to other orcs. lusite. Cryptomelane and delta samples gave a variety 6. The shape of the fine fractions as seen in th e of types of heating curves. It is understandable electron microscope is distinctive. that this would be true when one considers the 7. The specific surface of commercial sample range of composition of samples with cryptome varies widely, being highest on artificial ox-ides. ' lane-type patterns and the fact that th ese samples are not pure. The first main temperature eff ect ranges from 550 0 C in sample 43 (fig. 9, b) to The authors thank the U. S. Geological Survey 760 0 C in sample 36 (fig. 9, d), and as a whole the and the commercial battery manufacturers for break is not as pronounced as with pyrolusite. furnishing samples of manganese dioxide. They From 700 0 to 1,100 0 C, the curves vary greatly. are also grateful to E. S. N ewman of tllis Bureau It becomes evid ent why this is so when the X-ray for help in obtaining the heating curves, and to data is examined. Som e cryptomelane samples, G. v\!. Vinal for many helpful suggestions. such as 43 and 36, go direcLly from cryptomelane to hausmannite with no intermediate stage of VIII . References bixbyite. Others ch ange gradwllly to bixbyite (as C1 and 48). This is probably due to the effect of [1] H. F . McMlll'die, Microscopic and diffraction studies certain impurities on th e transformation. In one on dry cell s and othel' raw m ateria ls, Trans. case (sample 26 ), a sample of pyrolusite gave a Electrochem. Soc. 86, 313 (1944). [2] W . le. Co le, A. D , Wad ley, and A. Walkly, An X-ray double break (flg. 8, f) between 600 0 and 700 0 C diCfractio.:l study of manganese dioxi de, Trans. and showed formation of h au mannite with only Electrochem. Soc. Preprint 12- 2. a parti.al fom aLion of bixbyite. orne of the ex Lra [3] A. Van Valke nburg, Jr. a nd H . F. McMurdie, High thermal eff ects at h igh tempera Lures were un ex·· temperat ure X-ray difl'raction apparatus, J . R e plained . Sample 43, for example, gave an un· sear ch NBS 38, 415 (1947) RP1782. [4] S. Speil , L 1-1. Berkelhamer, J. A. Pask , and B. Davies, iden tified phase above 1,200 0 C. Fusion causes Di fferentia l the rmal anal.v' is-it appli cation to some of the extra breaks at high temperat,ul'e in clays and other aluminous minera ls, U. S. Bur. cases of eertain samples. Mines, T ech. Pap. 664. [5] E. S. Newm a n and L. S. vVell s, E ffect of some added material on dicalci um s ili cate, J. R esearch NBS VII. Conelusions 36, 137 (J 946) RP1696. [6] A. Ferrari, L'esam e rongtenographico , dei reticoli 1. There are fiv e major types of manganese cristallini del fluoruro manganoso e del biossido di I dioxide of importance to battery teclmology: (a) manganese, Acad. naz. Lince i 3, 224 (1926). Pyrolusite; (b) gamma :MnO z, a form of pyrolusite [7] M. Fleischer a nd W. E. Richmond, The oxides of of fine crystal size, (c) ramsdellite, (d) crypto manganese, Econ. Geo. 38, 269 (1943). [8] O. Glemser , Uber eine neue 1IodificaLion des Mangan melane, a compound of variable compositi.on with dioxyds, Ber deu t. chem, Ges. [B], 72, 1879 (1939). , essential K or Na, (e) delta .Mn02, apparently a [9] F . Schossberger, Uber die rontgenograph U ntersuch finely crystalline form of cryptomelane. ung von naturlichen un ku nstlichen Braunstein, 2. The heating curves of the various types fall Physik. Ber. 22, 1340 (1941). into two general groups, one containing pyrolusite, [10] A. St. John, The crystal structure of manganese di oxide, Phys. R ev. 21, 389 (1923). ramsdellite and gamma, the other, cryptomelane [11] J. W, Grune r, Groll tite, H MnO" a raw mineral of the and delta. diasporegoethi te group, Am. J\[in eraI32, 654 (1947). 3. The latter group gives a greater variety of [12] J. W. Gru ner, The chemical relation hip of c rypto heating curvcs, partly because of a greater range m elane (psilomelane) h oll andite and coronadite, of composition. Am. Mineral 28, 497 (19'13). [13] C. Frondel and E. W . Heinrich, New data on hetaero 4. Impurities, which mayor may not affect the li te hydrohetaerolite, coronadite a nd hollandite oxides for battery use, have a strong influence on Am . Mineral 27, 48 (1942). the heating curves, which makes it doubtful if this [14] W. E. Richmond and M . F leischer, Cryptomelane, a
Modifications of Manganese Dioxide 599 new name for t he commonest of the " Psilomelane" the differential thermal analysis of clays, Compt. minerals, Am. Mineral 27, 607 (l g42). rend. 223, 383 (1946). [15] L . S. Ramsdell, The unit cell of cryptomelane, Am. [23] R. E. Grim and R. A. Rowland, Differenktl thermal Mineral 27, 611 (1942). analyses of clay minerals a nd other hydrous ma [16] C. Palache, H. Berman, and C. Frondel, The system terials, Am. Mineral 27, 746, 801 (1942). of mineralogy, 7th ed. (John ·Wiley and Sons, Inc., [24] R. .E. Grim, Different ial thermal curves of prepared New York, N. Y., ]935). mixtures of clay mineral s, Am. Mineral 32, 493 [17] W. Feitknecht and W. Marti, Uber die Oxydation, (1947). von Mangen hydroxyd mit molekalaren Sauer [25] R. W . G. Wyckoff, The structure of crystals, 2d ed., stoff, Hel. Chern . Acta 28, 129 (1945). p. 290 (Chemical Catalog Co., Kew York, N. Y. , [18] M. P . D uBois, Contribution a l'etude des oxyden du 1931). Manganese, Ann. de Chimie 5, 411 (1936). [26] M. H . J elJinek and 1. Fankuchen, X-ray diffraction [19] Von Julius Meyer and R. Kanters, :lur Kenntnis des examination of gamma alumina, Ind. Eng. Chem. dreiwertagen Mangans VII, Z. anorg. Chem. 185, 37, 158 (1945). 178 (1929). [27] G. H. Cameron and A. L. P atterson, Symposium on [20] H. B. Weiser, Inorganic colloid chemistry, II, t he radiogr'1phy and X-ray diffraction, Am. Soc. hydrous oxides and hydroxides (John Wiley and Testing Materials, p. 324 (1936). Sons, Inc., New York, N. Y., 1935). [28] S. Brunauer, P . H . Emmett, and E. Teller, Absorp [2 1] P . H. Emmett, Symposium on new methods for tion of gases in multimolecular layers, J. Am. particle size determination in the subuene range, Chern. Soc. 60, 309 (I 938). Am. Soc. T estin g Materials (March 1941). [22] S. Cailiere, S. H enin, and L. Tane, Investigat ions of "WASHINGTON, June 11, 1948.
600 Journal of Research