Summary of Synthetic Asbestos Research
HajimeSAITO
INTRODUCTION
Asbestos has been used as various industrial material， chiefly as heat and electric insulating material， and electrolytic diaphragm． Generally， asbestos can be classifed into two groupeg though there are various types． One of them is serpentine group and the other is amphibole group． Asbestos used as indust ial rn郎erial are chrysotile belonging to the former group， and amosite and crocidolite belonging to the later group． In those asbestos， chrysotile， amosite and crocidolite have been chifly used as electrolytic diaphragm， heat insulating material and acid proof material in obedience to their properties respectively． In this investigation， the fundamental studies were systematica！ly done to prepare f1exible fluor−synthetic asbe3tos having length of fl’oM 3／8 to 3／4 inch．HISTORICAL SURVEY AND PRESENT STATUS
a．Synthesis of Serpentine Group It has been considered that one−fourth of combined water and the others in crysotile crystal exist in the form of molecule and ion respectively． Accordingly， synthesis of crysotile has been done by hydrothermal method． It was reported、that the fine crysotile crystal was obtained under pressure by Ipatieff and Mourom＄tSeffi）from silica gel and magnesium salt solution． by Syromyatrukow2）from silica gel and magnesium hydroxide solution， by Wells3）from sodium silicate and magnesium carbonate solution， and by Jander and Wuhrer4）， and∫ander and Fett”・） from the fundamental studies of MgO−SiO。．−HL，O system． Finder and Gruner6）obtaine5 crsotile by the treatment of olivine with sodium sulphide solution under presSure．．The research on MgO− SiO2−H20 system has been done by Branderger， Epprecht and Niggli7）， Eprecht8）， Balduzzi， EpPrecht and Nigglig）， Nolllo， Bowen and Tutte111）， and Roy and Roy12）etc．， It was reported by Balduzzi that chrysotile single crystlal having length of O 15 cm、 was obtained． Bauer and Keat13） synthesized crysotile under higher temperature and pressure than the conditions of the former reSearcheS． As above mentioned， The synthesis of cry30tile hεs been done by hydrothermal method． As a result， crystals obtained were very small in every cases． Experiments on the crystallization by fusion method from the batch composition substituted OH ion for F ion in crysotile crystal lattice were done by Noda， Sugiyama and Otsukai4）． It was reported that crystals belonging to serpentine group instead of crysotile were always obtained． Recently， it was considered that each and every combined water included in crysotile exist in the form of OH ion． Therefore， synthesis of crysotile may be possible to prepare by fusion method by substituting OH ion for F ion． The researches were done from this point of view． b．Syntheis of Amphibole Group Many researches on synthesis of amphibole group have been done by fusion method from th6 batch cornposit量on substituted OH ion for F ion in the composition of amphibole for the reason that互ll combined water in it exist in the form of hydroxyl ion． Ltidkeis）synthesized amphibg】e5
Jun． 1958． Reports of the Faculty of Engineering．Yamanashi University No．9 ・・y・t・1having 1・ngth・f f・・m O・1 mm・t・0・3 mm・fr・m鋤t b・坤・B・w・・aud S・h・i・er16）・ Grigoriev and Iskiill’i7）， and Hirose18）also reported the crystallization of amphibole from the melt including fluoride． Recently， Comeforo and Eite119）， and Comeforo and Kohn20）synthesized various kinds of fluor−amphibole by fusion method and reported on the detailed investigation regarding to the obtained crysta1． The author have been doing synthetic fluor amphibole research since 1943． Experiments on the cryst．allization of fluor−amphiboles from melts were carried on in the course of research on the synthesis of asbestiform minerals． The lnass obtained was consist of the glassy substances including asbestos fibers． Therefore， the crystal separated from theまnass by hand sorting is always brittle because it has glassy substances between the cleavage of crysta1． It cannot only be utilized as the use of spinning a川weaving but it is freqUently destroyed on the occasion of separating from the mass by hand sorting． Accordingly， the research on hydrothermal treatエnent was carried on for the purpose of separating fine and flexibe crystal from the mass．
EXPERIMENTS
1．Experiment on Crysta1】ization of Fluor−Amphibole from Mel V21） Experiments on the crystallization of fiuor−amphiboles from melts Were carried on in the course of research on the synthesis of asbestiform minerals． Among amphibole asbestos， natural riebeckite（blue asbestos？）is known to be most flexible and able to spin． No crystals separated輌n the mass obtained from the melt of the composition 皿 1ff corresponding fluor−riebeckite （FちNa2Fe3 Fe2 （Si4011）2） by cool ing at the rate of 300C per hour． Therefore， batches of vaτious compositions substitut．ing．・ferroug ion in fluor−riebeckite composition with other crystallo−chemical ly equivalent ions tested． Welldeveloped needle crystals sgparated from melts containing magnesium ion instead of ferrous ion． The best result was dbtained with the batch in which excess amount of basic component3 were used and a part of the sodium ion was replaced with calcium ion． The mass consisting of large parts of crystals of 1・5−2 mm．10ng and O．1 mm． thick was resulted from the melt of the composition of 2 NaF・CaO・3MgO・ Fe203・8SiO2 in a 30g scale experiment． ・・．・ Some of these crystals were separated from the mass by hand soring for optical measurements． The crystals Were confirmed to be a kind of soda−amphiboles from the optical data 6hown in the followingS： Angle betwcen cleavage faces 1240 Z〈C 15−20° 2V 70−800 Birifringence O．025−0．035 2．Suitable Amount Range of Alkalioxide in Raw Materials Mixture for Obtaining Fine Crystals？2） In the first e ．K periments on the crystallization of fluor−amphibole from melts， the best result was obtained with the batch in which excess amounts of basic components were used and a part of the sodium ion was replaced with calcium』ion． As the amount of basic component and the ratio of calcium ion to sodium iQn in the batch increased， the crystallization was mっre easily， but the crystalline part in the mass became graduglly spatiller and the cryst41s became more brittle． In those experiments， the triangul5r diagraln， where the the following three co皿ponents occupied each apex． thfit is，2Na20・2MgO・MgF2・Fe203・8SiO2；2CaO・2MgO・MgF2rFe203・8SiO2；and 2MgO・MgFi・Fe203・8SiO2， was used． Thu§， each point in that diagram could express tho wholeSu皿mary of Synthetic Agbestos Research co甲position of the batch， changing the total amount of basic copapOぬents and the ratio、of calcium ion lto sodium ion in them． The masses were obtained from the melts of the composit’ions expressed with each point in that diagram by the Quenching Method・The suitable amount range of basic components and the suitable ratio of calcium oxide to sodium oxide in the batch were studied from the size and the amount of crystals in each mass by microscopic observation． As a resu五ts， fhe suitable tota1・amount range of alkalioxide wa8 found to be from 1．50 to 1．65 times as much as the theoretical amount． The suitable ratio range of calcium oxide to sodium oxide was from O・25 to O・50． At the same「time， the temperature of crystallization and that for obtaining the largest crystal for each composition of batch， was determined． The process of 、c rystall izat ion was studied by the measurernent of refractive indices of glassy substances in each mass at each experimental temperature． 3．Suitable Amount Range of Fluorine in B亘tch for Prevention of Formatio・n of Pyroxene and Olivine in Crystallized Amphibole Block23） In the egperiments on the crystallization of fluor−amphibole from melts， foreign minerals， such as pyroxene and olivine were separated out together with fluor−amphibole．in the mass obtained from the composition having the theoretical amount of fluorine， even if the vaporization of fluorine was prevented． In order to determine the necessary amount of excess fluorine to prevent the formation of pyroxene or olivine minerals， the following experiments were made． As it was observed in the previous experiment that the excess of basic components in the melt resulted good needle crystals， we used in the present experiments the composition of 1・25Na20・o・40cao・ 2MgO・MgF2・Fe L・ 03・8SiO2 as a standard composition and varied the ratio of MgO：’ lgF2 ＝・n：m in the general forrnula 1．25Na20・0．40CaO・2nMgO・1nMgF2・Fe203・8SiO2． This general formula ユnay be rewritten as followsl 1．25Na20・040CaO・2nMgO・mMgF2・Fe203・8SiO2＝n（1．25Na20・0．40CaO・2MgO・MgFp．・Fe203・ 8SiO2）十（1−n） （1・25Na20・0・40CaO・Fe203・8SiO2）十（m−n）MgF2 That is to say that each composition may be represented in a point in a composition triangular where following three components occupy each apex 1、25Na20・0．40CaO・2MgO・MgF2・Fe203・8SiO2； ユ．25NaL・ O・0．40CtiO・Fe2 O・3・8SiO2；and MgF2． This sections for the microscopic observation were made from quenched spe：imens in order to examine the formation of foreign minerals． Refractive nidices of glass in the quenched specimens were measured in order to follow the process of crystallization． As the fluorine content of batch increased， foreign minerals existing with fluor−− amph ibole crystals decreased， but the crystal growth of fluor−amphibole became more difficult． It was concluded that the suitable range of fluorine content for synthesizing good crystats of fluor−amphibole was in the composition． triangle having the following three apexes・n＝・1・00，］m＝1・00； n＝1．00，m＝1．18．；n＝0．82．1n＝0．95． The compostiion corresponding to the center of that riangle 、 has 1．05 times as much fl uorine as that of the ideal chemical formula． 4．Change of Chemical Composition of Batch due to Vaporization of Fluorine from Melt24） In the experiments on the crystallization of f1、uor−amphibole．、from melts， it was found that the composition of the melt changed from that of the batch because of containing volatjle fluorine and basic components． Therefore， the separation of foreign minerals from the melt such as pyro．tene and olivin， might be promoted． In the present experiments， we used several batch compositions which were m蚕de by fixing the value n and m， such as m plus n equal to one， in the、 general formula of nNa20・mCaO・2MgO・ MgF2・Fe203・8SiO2． Firstly， the general aspect of vaporization of fluorine for each compositiQn
7
Jun． 1958． R・P・rts°f th⊇・9・lty°f E・ginee「i・g Yam・na・hi U・ive「sity No．．9 was st・di・d f・・m w・ight二1…curve at thet・mperatu・el．・a・gOf f・・m・・Om．t・血P・r・tu・e t。 1250°C．As a reSult， the difference in theヅocesφ℃f vaporiZ’atiori of fluorine which reliu1 ted fro】［ri the diffetence of c6mposition was caused to some extent by the difference of the reaction mechanism and that of the temperature coefficient of each melt etc．， but the vaporized amount of fluorine waS found approximately const ant in each composition． Secondly， the form of vaporized fluorine was studied from chemical analysis of the mass obtained from such compositions as n and m equal to O5 in the general formula． It was concluded that fluor−fifths of the total amount of vaporized fluorine were vaporized in the form of SiF4 and one−fifth was vapo亘zed alone or with alkali component in tlle batch． Thirdly， the amount of vaporized fluorine was studied from chemical analysis of the masses obtained from various total amQunts of batch up to IKg weight with various sizes of crucibles having the diameters up to 15cm． It apPeared that the vaporized fluorine percentage was nearly constant in such experimental conditions when the ratio of the am凹nt of batch（g）to the square of diameter of crucible（cm）and the thermal conditions of experiments were constant． It was found from the diagram showing the relation between that ratio and the vaporized fluorine percentage that the vaporized fluorine percentage was comparatively smalI and approached a constant value when that ratio was over seven． In order to synthesise good crystals， the amounts of vapor’ized fluorine and silicon were only added excessively in the batch because in practice synthe8is was carried on with the composition of the batch having excess amo岨ts of basic component to the theoretical amount． It was found that the suitablee excess amounts corresponded to between five and ten percent of the theoretical amount of fluoine in accordance with the reaction conditions． 5．Influence of Reduction of Ferric Oxide in Raw Materials Mixture upon Crystallization from Melt25） In the gxperiments on the crystallization of fl uor− amphibole from melts・ a part of ferric oxide in raw materials mixture was reduced to rnetalic iron and ferrous oxide when a graphite crucible was used． Accordingly， it was considered that the，composition of me！t considerably was changed from the batch composition at the temperature of cr ystal deposit． The influences of this change upon the crystallization from melts were studied． Firstly， the batch composition used in this experiment was （1）J，2SNa：0・0．40CaO・1．88MgO・1．12MgF2・Fe203・8．06SiO2． After many two hundred grams of raw materials mixture were melted in graphite crucible at 1250°Cfor one hour， they were cooled sl owly till 1050°C at the rate of］2°C per hour．The contents of ferric and ferrous oxide in each masses were analised chemically and the following result was obtained． Under th．oge experimental conditions， fifteen percent of the total amount of ferric oxide in batch was reduced to rnetalic iron and fifty five percent of the residual ferric ・oxide was reduced to ferrous oxide． Secondly， the masses were obtained under the same experimental conditions from the following batch compositions． （2）050N a20・1．50NaF・O、40CaF￡・・290Fe203・8．05S iO2 （3）125Na20・0．40CaO・1．17MgF2・1．83MgO・1，95Fe203・8．05SiO2 （4）1．25Na20・0．40CaO・1．17MgF2・・O．53MgO・1．95Fe2 0 3・8．05SiO2 （5）1．25Na20・0．40CaO・ユ．17MgF2・0．53MgO・Fe203・8．05SiO2 ． ’ （6）1．25Na：C・0．40CaO・117MgF2・1．23MgO・FeiO2・8．05SiO2 Thin leaves of each masseg were prepared for microscopic observation of crystallization from melts．
Summary of Synthetic Asbestos ReseaτCh It・was proper that the best・result wa§obtained from the batch composition which supplied the shortage of the amount of ferric oxide produced by the reduction of ferric oxide and decreased the amount of magnesium oxide corresponding to the amount of ferrous oxide produced in the general batch composition（1）． The good result was obtained from the batch composition insafficient of the amount of ferric oxide and having the excess amounts of magnesium and ferrous oxide． Then， it was collsidered that the excess amounts of magnesium and ferrous oxide supplied the． shortage of the amount of ferric oxide． Therefore， the ratio of calcium ion to sodium ion in crystal was l arger than the other cases． Because the results obtained from the batch compositions， containing excess amounts of magnesium and ferrous ions or insufficient amounts of ferric oxide were considerably inferior than that from the above batch composition． 6．Effect of Various Additions to Raw Materials Mixture on Crystallization from Melt26） In the experiments on the crystallization from mel ts， various kinds of addition have be籠 frequently used in order to promote crystal growth． In those experiments， the batch compositions used were as follows： A．1．25Na20・0．40CaO・1．25MgO・1．75MgF2・Fe203・8．37SiO2 ・ B．0．50Na20・0．50CaO・125MgO・1．75MgF2・Fe203・8．37SiO2 The additions mixed with those batch compositions were NaHSO4， KHSO4， NaC1， NaH2 PO4・H20． Na2S203 etc． and the amount of them was ten percent of the amo狐nt of raw materials mixture． Such thermo−experimental conditions such as most convenient for the comparison of crystallization・、 for each batches were decided after experiments were carried on tmder various reaction condit−・ ions． After five grams of each raw materials mixture were melted at 1250°C for one hour for A− composition and at 1340°C for 30 minutes for B composition， the melt were cooled at the rate of 40°Cper minute for the former composition and at the rate of 10°C per minute for the latter1’ one． Thin leaves of each experiments wdre prepared for microscopic observation and the crysta− 11ization of each masses was compared． As a result， the effect of addition for A composition was hardly found for the reason that the，． excess amount of alkali oxide in raw materials lnixture compared with the theoretical coxnposition was not only effective as flux but its effect was extremely larger than that of addition． The effect of addition for B composition was remarkably found in the case of ・ add ition of five percentz of NaHSO4 and two percent bf Na2 S203 for total amount of batch to raw materials mixture． No good result was obtained in the case of addition of more excess amount of them to raw materials・ mixture on account of increasing the glass substance in the mass． 7．Studles on Suitable Fluoride Raw Materia127） In the experiments on the crystillization from melts， the volatile fluoride in raw materials． mixture was vaporized in the heating process of batch from lnelt． This proceeded thc deposit of” foriegn minerals from melt and、gave bad effect、on the crystallization of． fluor−amphibole． Therefore， it was necessary to manipulate the amount of vaporized fluorine to raw materials・ mixtu士e in order to obtain good result． But it was difficult to manipulate it completely as． much as possible for the reason that its amount varied、considerably with the change of experi− 1℃ental conditions 1 when the crystal liza．t ion from melt was ． performed from部ch batch composition that very large amount of flu6rine was vaporized． Accordingly， it was desirable to select stable． fluoride in the heating Process of batch． ・9
Jun． 1958． Reports of・the FacUlty．df Eligineering Yamanashi． Unversity No．9 In those experiments， the raw material8 mixture containing MgF2’NaF， CaF2． and Na2 SiF6 as f1凹oride．were prepared． For example， the batch co血position in the case of using MgFg as fluorine 、as fo110ws： … ’ 1．20Na20・0．50CaO・1．75MgO・1．25MgF2・Fe203・8．10SiOヵ After sixty five grams of each raw置naterials mixture wa■melted at 1260°C for one ho叫it was ・cooled at the rate・of 12°C per hour till 10500C． Thin leaves were prepared for the study of ’crystallization from each melt and the fluorine content of each masses was determined chemicall．y by the therium nitrate method． Froln the qxperimental data， it was mbst desirabl e to select CaF2 as fluoride raw rnaterial， but the obtained crystal8 from the raw materials mixture contain4ng total 、amount、 of fluorine as CaF2 were comparativel y brittle． Therefore， it was most suitable to select ．NaF or MgF2 as fluoride raw material by all means． ．8．Relation of Velocity of Cry8tal Growth with Chemical Composition of Melt28） In the experiments on the crystallization of fIuor−amphiboles from mel ts， it was important to study the verocity of crystal growth in order to decide the suitable cooling velocity of m61ts． In 『the present experiment9， the velocity of cry8tal growth was studied by tho Queuching Method． We used 8everal batch compositions which were made by fixing the value m and n in such range as m plusnwas equal toava1 ue between l and2， and the ratio nto m was equal to a va1．ue 『：between 43 and O．52 in the general formula of nNa20・mCaO・2MgO・MgF2・Fe203・8SiO2． ．Thin sect ions were made from quenched 8pecimens in order to measure the size of crystals by micrscopic observation． Those crystals grew in the melt for a certain n皿1ber of minutes， kept at the neighbourhood of the temperature of crystallization． The velocity of cry8tal grOwth was decided for each composition from the diagram showing the relation between the size of the largest cry8tal in each specimens and time after the melt wa8 kept at a defiロite temperature． As a result， the velocity of crystal growth was very different according to the difference of the ・composition of the batch， and varied from between O．OSmm and 2mm per hour． Whe担n Plus n was equal to 2， it was fourty times as speedy as when it was equal to 1． It was natural that the ・velocity of crystal growth becameエnore speedy as the ratio of n to m became larger， but the ’difference was general ly sma11．In the neighbourhood of one of the most suitable batch compositi、ons −such． as 1．20Na20・050CaO・2MgO・MgF2・F6203・8SiO2， it was lmm per hour． At thc same time， it was guessed that from the measurement of refractive indices of glass in the quenched specimeng， the ratio of CaO to Na20’in the obtained cry就al was always largeτthan that in the batch． It was understood over again that well deve10ped needle crystals could be separated fronl the melt ．with a batch composition in which a part of sodium ion in the theoretical formula was replaced with calcium ion as it was mentioned in the first paper． But of cour8e， the crystal obtained from the batch in which a mol ratio of CaO to Na20 was beyond one could not have well developed ・cleavage and were brittle． 9．Effect of Temperature Gradient produced in Mass on Crystallzation from Melt29） In th6 experiments on the crystaUization from melts， the electric fロmace ha▼ing uniform temporature distribution has been usually used． Accordingly， the crystals separated from melt grew comparatively at random and it was difficult to obtain large single℃rystal for the reason ’that such separated crystals had large sectional area． Iri tho8e experiments， the masses were obtained using the furnaCe having various temperature gradients with the same batch comppsition and reaction conditions． The experimeats were carried on・ill the temperature gradient range of from zero to 100°C per 4．4cm．height・Two hundr6d grams
Surnmary of Synthetic A8bestos ReEPearch of、batches were charged in the graphite crucible intered in crucible’keeping magnesia．in it in order to get uniform temperature gradient and to’meagure the correct reaction temperature． The electric furnace being fitted with siliconit−heizrohre was used in those experiments． It wag shown ．in Fig．1． The batch composition was as follows： 1．25Na20・0．40CaO・1．85MgO・1．75MgF2・Fe203・8．04SiO2 After each batch wa8 melted at 1250°C foc one hour， it was cooled at the rate of 12 r 15°C per ．、hour． The derection of crystal growth alld the 8tate of crystal deposit in each mass were studied from the observation by the naked eye and their thin leaves． As a result， it was found that it i8 necessary to choose the proper range of temperature ．gradient in order to obtain the regUl arity of c町8tal growth． The best result was obtained with 狽??@temperature gradient of 15°C per lcm． height of melt． ユ0・Experiment on Separation of Crystals from Crystallized Mass by Hydrothermal Treattnents30），31） The mas8 consisting of large parts of crystals of 5−10m血10ng was resulted with a graphit6 crucible from the melt of the composition of 2NaF・CaO・3MgO・Fe203・8SiO2 in a lkg scale ・experiment by cooling at the rate of 12−15°C per hour． The mass obtained was shown in Fig．2． On the occasion of separating thQse cry8tals from the masg by hand sorting， crystals were 』frequently destroyed and the glagsy substances’existing between the cleavage of crysta18 could ・not be relnoved． Therfore， the mass was treated with water， and with various concentrations of various aoid， alkal i and salt solutions at 300−3500C under prossure． In those experiments， the crystals themselves w6re sometimes destroyed， whe血the 8evere react ion conditions were chosen， and the crystal s could not be separated thoroughly from gla8sy substallces， when the mild reaction ・conditions were chosen． Accordingly， the 8uitable kindS of solut ion of treatment and the suitable モ盾獅р奄狽奄盾獅刀@of reaction were studied． For example， well developed needle−shaped crystals of 10−15microns in width could be separated from the mass by the treatment of 2N NaOH or Na2 CO3 sqlution for 8−16 hours at 350。C． SOme of those crygtals obtained were flexible to a certain extent compared with crygtals separated from the mass by hand sorting． In those experiments， a part of the fluorine ion eXisting at「 狽??@crystal lattice points was replaced with hydroxyl ion． The degree of substitution and the thermal stability of the cry8tals were studied bゾweight−10ss curVe． In the experimenSs°n the c・「ystqlli・qti°n・f flu°・イi・beckiC・f「・m m・lt・・th・b・・t． 「esUlt wg・ obtained with the batch in which a part of sodium ion was replaced with calcium ion． But as the amoqnt ot calcium ion increased， crystals obta量ned becalne gradually more brittle． Therefore， in those experiments， the masg was resulted from the melt of the composition of 2NaF・3MgO・Fe203・ 8SiO2 in a lkg scale exprimeht by cooling at the rate of 12−15°C per hour wi傲h a graphite crucible．、 In the present experiments， the influence of reacting time and concentration of solutions upon hydrotherma1・reactions was 8tudied using the typical acid，．alkal i， and galt solutions for obtaining comparatively fine crystals ’ which were usedでin the previous experiment． After the crystalswere ・separated・from the mass・by．hydrotherma1・t reatments under pressure． The changes of their thermal stability and their optical properties were studied by the weigh炉10ss curve and with・ oil immersion technique． FK）m tho8e data， it wag囎derstood・that a part of the fluorine ion in the cry帥al was・repl．aced with hydroxyl ion and their thermal stability wa8 inferior to lmtrea‡ed crystals． It appeared that it was】more difficult to obtain fine ・crystal by−hydrothermal treatme勤t from the mas9、 obtain6d・from th●melt of、the above than from the mas80btain⑤d from the m61t
11
JUI1． 1958． Reports，of the Fac url−ty ’ of Engineering Yamanashi Unversity No．9 of…the ．composition．inlwhich ；a part of sodium ion was replac6d with calcium ion、’For example， well developed needle−shaped crystals of 10−15 microns in width could be separated fr（m the mass，by the treatment of l N HC1、for 8−16 houτs， and 4N NaOH for 24．hours at 350°C． In those 6xperiments， the process of separation of fine crystal by hydrothermal treatment was l shown in Fig．3． 11、Chemical Composition， Thermal Stability， Optical Properties， X−Ray and EIectron Microscopic Obsとrvation of宜ydrothetmally Treated Fluor−Amphibol e32） The masses con8isting of large parts of crystals of 10−15mm long were obtained by cooling atl the rate of 12−15°C per hour from the melts of the followimg compositions， i e．（a）1．OONa20・ o，65cao・2Mgo・MgF2・Fe203・ssie2；and 1．25Na20・o．40cao・1．88Mgo・1．12MgF2・Fe203・8s io2’（b）． The・： masses treated with 2N Na2CO3 and N／4 HCI at 350°C皿der pressure． Those compositions of batch and the kinds of solution of treatment were found to work well from the results of the・ previous experiments． Fine crystals of 10−15 microns thick were separated from each mass． Their’ chemi．cal compositions were determined by the chemical analysis・and theirρptical properties』 aud thermal stability were measured・Also・X−ray and electron microscopic observations gf them were carried．on． As a result， the chemmical compositions of synthetic fluor amphiboles obtained． were found to be （・）（0・．・50H・．13F1．・・）Na・66C・・．・・Mg・．・3F・丑1F・認、（Si・．・90・・）・and （b）（OH・．13F・，87）N・・．s4C・・，4・Mg・．92Fg。1，、F・語、（Si・．・・0・・）・・ The optical constants were mea8ured with a peterographic microscope using the oil immersio廿・ tech血ique as follows： （a） （b）
α：1．600−1．608
1．597−1．599 γ一α ： 0．013−0．015 0．021−0．022 Z〈C ： 19−200・ 160 1t seems that under those experimental conditions， the amount of replacement of fluorine ion by・ hydroxyl ion in treated fluor−ampllibole is smaller than ten percent of the total amount of’ fluorine， and thermal stability was not harmed much except for the one treated with hydrochloric acid． The needle minerals synthe＄ized were confirmed to be fluor−amphibole with X−ray diffra− ction data， too． The X−ray examination also revealed that the fluor−amphibole minerals could no七 be changed to other minerals by the hydrothermal treatments under the conditions described above． It was observed under an electronmicroscope that needle crystals of fluor・−amphibole had more・ or less imperfect cleavages and had been broken into short fibers or even corroded by hydrothe− rmal treatments． The electronmicroscopic observation was shown in Fig．4． 12．Crystallization from Melt of Batch Composition System MgO−−MgF2−SiO233） The suitable batch composition was studied in qrder to obtain the good crystallization from・melt． of MgO−MgF2−SiO2 system． Sodium carbonate was used as flux．for the reason that the melting： point of this system was very high． In each experiments， thirty grams of batch was melted at 1250°Cfor half an hour and cooled at the rate of 300C per hour． The suitEble batch composition・ was decided by、microscopic observation of thin leaves of each mass． The chosen standard comp− osition was 2MgO・4MgF2・4SiOi． The．crystallization from melts was carried on from the batch compositions which were added sodium carbonate in the．range of．from Qne、to six mol to the・12
Summary of Synth．etio Asbestos Research standard co血position and which were varied the alnount of magnesium fluor．ide in the range of from five to two mol because the forign minerals were separated owing to the vapori8ation of flu・・i…．・・d whi・h w・・e va・i・d the・m・unt・f・ili・a i・th。・ang。。ff・。mf・。，t。、i。m。1 b。、a．，e apart of silica was vaporised with fluorine from melt， and which included a part of fluorine as sodium fluoride because sodium fluoride was more stable th亘n magnesium fluoride． The maSs containing many crystals having length of 5mm was obtained from the batch compr osition of 3MgO・3MgF2・5SiO2・2NaF・Na3CO3・Crystals separated from the mass by hand sorting wers confirmed as rhombic hexagonal amphibole from X−ray diffraction data． Their optical prQp− erties were determined by microscopic measurements as follows： Refractive Indices：α1．537−1．541γ1．550−1．554 Birifringence ： 0．615−0．020 Z〈C ： 5−10°’㌦、
CONCLUSIONS
1．CrystaUizatlon of Fluor−Amphibole from Melt It wa8 concluded that the following reaction conditions were necegsary to obtain the mass consisting of large parts of long single crystals having well developed cleavage． The batch composition i・e・1・25Na2 O・0．40 CaO・1．15MgF自・0．05MgO・1．95Fe203・8．05SiO20r mixed NaF as fluoride raw material was prepared． After the batch was perfectly melted in a graphite crucible having a constant size decided from the quantity of batch in the electric furnace having the temperature gradient of 15°C per lcm・heighちthe melt was slowly cooled with the speed of O・5°Cper hour・Thus・the good mass c藺taining long crystals having length of more 3／8 inch wa・・bt・ined because the c・y・t・l g・・wth fiロi動・d i・t，，h・t・mp・・ature rang・・f fr・m th・t・mpera− ture of crystal deposit to under 5°C of it and the velocity of the crystal growth was lmm per hOur． 2．Separation of Fine Crystal from Mass It was concludeまthat the good result wag obtai血ed by means of the following chemical treatment to separate fine crystal from the mass． When the mass was treated hydrotherlnally with acid， alkali or salt solution under the suitable reaction conditions， fine crystal of under ten microns in width could be separated from the mass． By that treatment， no crystal was destroyed and the glassy substances existing between the cleavage of crystal could be excellently removed， th・ugh it wag f・equently dest・・y・d・nd th・y・・u！d n・t be rem・v・d・n th・・cca・i・n・f・ep・・ating crystal from the mass by hand sorting． The suitable reaction conditions for the treatments were solnewhat different with the difference of composition of the mass． They were also different with the difference of the purpose of its practical use． It was proper that the mass was treated with from two to four normal sodium hydrate・’solution（high concentration of alkali solution）or from a half to one normal hydrochloric acid．（10w concentration of acid），when▲t was required t・get fine crystal separat・d fr・m the mass having’ 狽??@thermal stability・r the flexibil ity suitable for spinning respectively． The difference of effect of those two solution o丘crystal was dne to the diffe「en・e・f th・・eacti・n m・・hani・m f・r glas・sub・t・nce． Th・・uit・b1・・ea・ti・n t。mperat。，e and tim・were f「・m 250°C t・350°C・・d f・・m 16 hrs・t・24h・・． respect量vely i・th・t・eatm・nt・f・・y・t・1 with two solutions． 3・Investigaton of Obtained Crystal タ ‘・ i3Jun． 1958． Reports of the Faculty of Engineering Yamanashi Unversity No．9 Fine crystals of 10−15 microns in width were separated from the mass by hydrethermal reatMent． Th・i・ghemical・・mp・siti・n and・藤ica1・・p牢命w鱒・measur・d・lt w・・pressum・d from the data of their chemical composition and weighト10ss curve that a part of the fl uorine ion exiSting at the crystal weight−10ss curve that a part of the fluorine ion exist in the crystal lattice points wag replaced with hydroxyl ion by hydrothermal treatment in those experimental conditions． Synthesis of fhlor−amLphibole was confirmed from the X−ray diffraction data and the optical constants of the obtained crystal． From those data， no change of crystal to other minerals was、also confirlned． It was concluded from th6 electronmicroscopic observation that the synthesis of mass must be carried on in’the reaction conditions of being able to obtain crystal hav ing the wel1−developed cleavage and including crystals in the mass in a constant direct ion， and only the grassy substan− ces existing between the cleavage of crystal did not effect on the flexibility of crysttl，．
REEERENCES
ユ）W．lpatieff and D．Mouromstseff， C・R・Acad・Soc・US・S R 185・647（1927） 2）F．V．Syromyatrukow， Econ． Geol・89，（1935） 3）F．G．WeUs， Amer．T．Sci．18．35（1929） 4）W．Jander aptd J．wqhrer， Z．Anorg・Allgem・Chem，235，273（1938） 5）W．Jander and RFett， Z．Anorg． Allgem・Chem，242・145（1939） 6）」．L．Linder and JW．Gruner， Econ・Geo1・旦｛，537（1939） 7）F．Brandenberger， W．EpPrech ahd P・Nig91i， Helv・Chim・Acta，30・9（1947） 8）W．EPrecht， Schweiz，Mineral・Petro・Mitt・鉦！1（1947） 9）F．Balduzzi， W．EpPrecht and P．Niggli， Schweiz Minera1・Petrog・Mitt・，8塑93（1951） 10）W．No11，Z．Anorg．Chem．，261，（11950）；Naturwiss，．3L7，540（1950） 11）N工．Bowen and O．F．Tuttle， Bull． Geo1・Soc・Am・，60，439（1949） 12）D．M．Roy and R．Roy， Am・MinraL，i坦ρ57（1954） 13）W．H．Bauer and P．P．Keat， Cerm． Age・，56，16（1950） 14）T．Noda， S．Sugiyama and K． Otsuka， J．Ceram． Assoc・Japan・巨麺9（1946） ユ5）W．Ltidk・， F・・tsch・． d． Mine・al・18，29（1933）；R・i・h・b…Ch・m・・玉121（1944），t 16）N．L．Bowen and J．F．Scheirer， Amer・Minera1・，20，543（19こ5） 17）D．P．Grigoriev and H．W．lskul 1， Zentralb・Minera1・，1936A・32 18）M．Hirose． Jap． Patent 161256（1944） 19）J．E．Comeforo and W．Eitel， Amer・MineraL，旦，368（1951） 20）J．E．Comeforo and J．A． Kohn， Amer・Mineral・，39，537（1954）；処．410（1955） 21）T．Noda and H．Saito， J． Ceram． Assoc・Japan，一．5＿4，，59（1946） 22）H．Saito， J．Chem． Soc： 23） 24） 25） 26） 27） 28） 29） 30） 31） 32） 33） do． do． do． 『 do． do． do． do． do、 do． do． do， Ind． Section Japan，55，703（1952） 56，585 （1953） 57，483（1954） 59，1309（1956） 59，1312（1956） 59，1309（1956） 57，536（1954） 59，1312（1956） 55，415 （1952） 56，235 （1953） 56，756 （1953） 59，1315（1956） 〔Received March 20，1958〕Fig．1 Fig．2 Fig．3．（X100） （1） （2） （3） （4） ： ： ： 存 （1） （2） Separated crystal by hand±sorting， average width of crystal O 5mm． Treated ’tcrysta 1 by hydrothermal treatment， do． 0．2mm． do． ， do． 15micron do． ， do． under 10micron
（3） t” Fig．4．（X3000） （1） 燕・， ． ． ’ 一撫、七@ 豫 翌
