the american mineralogist, vol. 53, may-june, 1968 · 2007. 4. 4. · the american mineralogist,...

THE AMERICAN MINERALOGIST, VOL. 53, MAY-JUNE, 1968

SILICATE APATITES AND OXYAPATITES

JUN Iro, National Bureau oJ Stand.ard,s,Washington, D.C.20231.

Ansrnacrvarious synthetic compounds of the silicate-apatite group are isomorphous with the

minerals abukumalite, beckelite, britholite and lessingite. Three different types .vere syn-thesized hydrothermally at 2 kilobars: (Ca, Sr, Ba, pb, Mn and Cd)ar+(Ln, y)us+SfuOrn_(OFI)r, (Na, Li)r+(Ln, Y)C+Si6Or4(OH)z and Naz+Ca"H(Ln, y)e_,3+SL_"p"Oza(OH)r, a:0to 6.

complete solid solution exists between hydroxylapatite caropoogr(oH): and ca+yesioOz(OH)z (abukumalite). Unit-cell dimensions of ail compounds and indexed X-ray powderdata of the typical end compounds are tabulated.

Anhydrous silicate oxyapatites, with the probable formulas Mz2+(Ln, y)sa+SLOzo andM+(Ln, Y)g3+SkOzo and more generally (M*, M*, Lna+, y3+ and Tha+)1s(Si, p and B)6Ct26,were prepared at temperatures from 950 to 1260oc in air. The following solid solutionsystems of oxyapatite were established: CazyrsiaOzedCasyzpoOzo, CazlaaSioOzsACaslaz_P6Oz6, Pbs4+Pbb,+Yrsi6or6aiPbz2+YsSioOrs, MgryrSiuO*eyroSirBzOgo, and CazyasioOze=YroSLBzOze.

rnfrared spectroscopic analysis, water determination by two difierent methods, anddifferential thermal analysis of selected compounds demonstrated the presence of oH insilicate-apatite and its absence in oxyapatite.

NalarSigoro was obtained as transparent needles up to 0.05 mm by slow cooling in aflux (NazWOr) from 1150 to 750oC. Euhedral transparent crystals of NaalneSioOzFz up to8 mm were grown by the combination of slow cooling and evaporation of the flux (NaF) attemperatures from 1350 to 900oc. New fluorescent yttrium and gadolinium analogues ofoxyapatites activated by europium and terbium were prepared.

INrnotucrroN

The isostructural relation of the minerals lessingite (zilbermintz, lg2g),britholite and beckelite (Morozewicz, l90s and winther, 1901), as wellas abukumalite (Hata, 1938) to apatite, calcium phosphate Caro(pOr)o(OH)r, hexagonal P6s/m with the general formula (Na, Ca, Th, Lnr16(Si, P)6Oz4(OH)2, was first pointed. out by Macharschki (1939) and wasconfirmed by later studies (Hagele and Machatschki, 1939; Gay, I9S7;omori and Hasegawa, 1953; Nechaeva and Borneman-Starynkevich,1956; Sakurai and Kato, 1962; Kupriyanova and Sidorenko, 1963).Tritomite (Neumann, Sverdrup and saebo, 1957) and spencite (Frondel,1961) are metamict minerals, which were probably fluoro-boro-sil icateapatites. They recryst allize as oxyapatite with the formula (Ln, Th, Ca)16_(Si, B)601402 when heated (Jaffe and Molinski, 1962).

The synthesis and crystal growth of analogous end members of theabove rare-earth silicate apatites have been an attractive field for re-search. Triimel and Eitel (1957) reported the growth of abukumalite

I Ln-Lanthanides.

SILICATE APATITES AND OXYAPATITES 891

CarYoSioOza(OH)2, and La and Nd-britholite (Na, Ca, La, Nd)roSieOr+-(OH), by solid-state reaction or flame fusion in air at high temperature.They speculated that hydroxyl ions were introduced from atmosphericmoisture. Ce-britholite (Ceftosil) (Ca, Ln)roSioOzrFz was found in blast-

furnace slags (Rudneva, Nikitin and Belov, 1962) and a similar

compound was synthesized by Lenov (1963). Keler, Godina and Sav-

chenko (1961) I Lenov (1962); and Miller and Rase (1964) synthesized in

air rare-earth orthosilicates having the apatite structure' with empirical

formula LnnSirOrz (Ln:La, Nd, Ce, Sm and Gd). At the time of prepara-

tion of this manuscript, Cockbain and Smith (1967) published the re-

sults of the syntheses of alkaline-earth rare-earth silicate and germanate

apatites by solid-state reaction at atmospheric pressure. They claimed

that hydroxyl compounds were synthesized simply by passing steam

through the reaction chamber at high temperature. They proposed a

classification of naturally occurring minerals based on their results from

rather l imited experiments. They also showed random distribution of

La3+ and Ca2+ over the 4/ and 6Z sites, and reported apatites with halogen

sites vacant, partly occupied or fully occupied, depending on cation

ratio.None of the above studies for hydroxylapatite shows proof of the

presence of OH ions in synthetic apatites. Unlike phosphate-apatite,sil icate-apatite loses water at temperatures below 1100'C. In my view to

be presented here hydrated end members can only be synthesized hy-

drothermally.Oxyapatite,l in which 2o sites are partly or completely vacant, has

been the subject of controversy since voelckerite Caro(PO+)oO was de-

scribed (Rogers, l9I2). It is certain that the difficulty of high tempera-

ture (1400'C) water analysis is responsible for earlier misleading reports.

Korber and Trdmel (1932) discredited the existence of oxygen-deficientoxyapatite Caro(PO+)eO previously reported (Trtimel, 1932); however,

they synthesized oxyapatite CaalazPoOgo with 26 oxygen sites fully oc-

cupied. Nevertheless, the concept of oxygen-deficient oxyapatite per-

sisted. Wondratschek (1963) described series of oxypyromorphite Pbro(PO4)6O, Dietzel and Paetsch (1956) synthesized a single crystal of lead

oxysilico phosphate apatite PbroSizPaOzr in which all the halogen sites

were said to be vacant. Recently Young and Munson (1966) reported a

partly dehydrated oxygen-deficient apatite from Colorado (OH, O, F

plus CI:25.68 compared with 26). None of the above studies, however,gave sufficient evidence to establish the presence of vacant halogen sites

I In this paper, compounds of the type CaroPrOz(OH)z will be referred to as hydroxyl-

apatite, those of the type CaroPoOzrOz as oxyapatites, and those of types CaroPoOzO as

oxygen-defi cient oxyapatites.

892 JUN ITO

(McConnell, 1965). After Kuzmin and Belov (1965) discarded the em-pirical formula for rare-earth orthosilicate apatite LtuSisOrr and intro-duced the cation defect formula Lna.66Si3O13, Cockbain and Smith(1967) again brought up the same concept of halogen-site vacancy.

My investigation was undertaken to establish isomorphous relationsbetween rare-earth sil icate and phosphate apatites, and to give additionalinformation for the characterization of these complex rare-earth silicatesby the determination of the conditions and the complete compositionalrange of formation. It was later extended to include the oxyapatiteproblem in an attempt to contribute to its clear definition, especially asregards the deficiency of sites and total charge balance. Moreover, newfluorescent l ight emitters activated by rare-earth elements were soughtamong these compounds. An investigation of single-crystal growth wasundertaken with sodium tungstate and sodium fluoride flux fusion inorder to supply good quality single crystals for further studies of the de-tailed structure of apatite and for possible applications in materialsscience.

SvNrnpsrs

Reagents used for the preparation of starting gels included oxides of lanthanum,neodymium, samarium, gadolinium, dysprosium, yttrium, erbium, holmium and lutetium;thorium nitrate, manganese chloride, strontium chloride, calcium nitrate, lead nitrate,magnesium nitrate, barium chloride, barium hydroxide, cadmium sulfate, lithium car-bonate, ammonium dihydrogen phosphate, boric acid, sodium silicate and potassium sili-cate.

Starting gels for lead rare-earth silicates were precipitated at pH 9 with ammoniafrom 300 ml of 0.5 millimolar solution of apatite. The resulting precipitates were washed ina centrifuge with 300 ml of distilled water. The gels for calcium analogues were preparedby precipitation with an ammonium carbonate solution stabilized in ammonia at pH 9.Gels for strontium, manganese and magnesium analogues were prepared by mixing Mg-(OH)r, MnCOa or SrCOs (all powdered) into a suspension rrf rare earth silicate at the timeof precipitation by ammonia at pH 9. Use of NaOH solution was avoided because analoguescontaining divalent ions readily form solid solutions with sodium analogues of the silicateapatite. The gels of the sodium, barium and lithium analogues were prepared in a porcelainevaporating dish of 100 ml capacity, by mlxing exact amounts of concentrated solutions ofsodium silicate, lithium carbonate or barium hydroxide with the precipitated rare-earthsilicate gels. Phosphate- and borate-bearing gels were also prepared in the same way asdescribed above, mixing solutions of boric acid, phosphoric acid or sodium phosphate intothe precip i tated s i l icate gels.

All the gels were air-dried at 1 10"C, except the manganese and barium analogues. Thesewere dried in an evacuated oven at 120"C in order to avoid oxidation or absorption of car-bon dioxide, respectively.

Dry syntheses were carried out by heating gels in a box furnace at temperatures rang-ing from 800 to 1250oC for 3 20 hours. A porcelain boat containing approximately 100 mgof dried gels was placed in a preheated furnace. Manganese analogues were synthesized at1250oC in order to avoid the formation of manganese oxides of higher valencies, whichmight be produced at temperatures below 1100oC.


Tllr,r 1 UNrr-Crlr DrurNsroNs or rnn Str-rcetr Ape'rrtes (uvonoxvr)

Synthesized Hydrothermally at 650'C and 2 kbars (450'C for Pb analogues)

893

Formula o (A) d (A) Y (4,) Formula ,, (A) c (A) Y (At

Mnala6 Si6Or4(OH), 9.66

MnrNdo Si6Or4(OH), 9.50

MnrSmoSioOzq(OH)z 9.43

MnrGdoSioOr(OH)z 9.38

MnrDyoSioOz(OH): 9.33

Mn+YoSioOz(OH): 9.31

MuEroSirOzr(OH): 9.28

carLaoSLOz(OH)z 9.66

CarCeoSiaOzr(OH): 9.61

Ca+NdsSioOa(OH)z 9.56

Ca+SmeSioOr+(OII)z 950

CarGdaSioOr(OH)z 9 43

CaupyoSioox(oH)z 9.40

CarYoSisOzr(OH)z 9 40

CarEroSLOzr(OH)z 9.38

Ca&uoSieOzr(OH)z 9.35

SrrLaoSiaOzr(OH), 9 7l

SrqCesSioO:r(OH), 9.67

Sr+NdaSioO:r(OH)z 9.62

SruSmoSisOr(OH)r 9.60

SrrGdsSLOs+(OH)z 9.53

SraDy6Si6Ora(OH)z 9.43

SrrYoSiaO:r(OH)z 9.52

SrnHooSisOzn(OH), 9.42

SrnEroSLO:r(OH)z 9.42

Sr+LusSiaOx(OH)z 9.50

PbaLaoSieOz(OH), 9.80

Pb4Ce6Si6o%(oH), 9.77

PbNdoSioOz(OH)2 9.76

Pb+SmoSioOa(OIl), 9 '74

PbrGdoSioO:q(OH)z 9.7 2

PbrD yaSisOz(OH): 9 . 70

Pb4Y6si6or4(oH), 9 68

Pb+EreSioOx(OH)z 9.68

PbrlusSioOzq(OI]), 9'(A

BanNdoSiaOza(OH), 9.76

BarSmoSiaOz+(OH)z 9.73

BarGdosieozr(oH)z 9.72

BaaDyoSirOx(OH)z 9.66

Na:YaSirOz(OH)." 9 34

NazLasSioOzr(OH), 9.74

7.05 s696.90 539

6.82 5256 78 52r6.68 5036.65 4996.63 494

7 . r2 5757 .09 5677.00 5546 94 5426.89 5316.83 5236 81 5196.78 5t76 .7+ 5107 .23 5917 . t4 5787 ro 5697.05 563

7.00 5506 .92 5326.9r 5436 .9 t 5316.83 5256.79 5327 .26 6047.19 s947 . r3 s887.05 s796 .99 5726.90 5626.86 5676.84 5516.75 543

7 . 1 87 . 1 07 .036 .95o . / 6

59J

582J / J

561512589

Hydrothermal syntheses were done in cold-seal steel bombs of 25 ml capacity at tem-

peratures from 500oC to 750oC under two kilobars pressure. Approximately 50 mg of dried

gels were placed in a silver liner made of 0.03 mm of silver foil (10 mm long and 3 mm LD.).

The bombs were filled with the exact amounts of distitled water to give the desired water

pressure. After the reaction, the bombs were always quenched by compressed air and the

the natural minerals in Table 4.

Fr-ux Gnowtrr or STNGLE CnYsrer,s

Single crystai growth of oxyapatite for crystal structure work and for chemical analysis

was attempted by flux fusion. An extended search showed that sodium tungstate was an

effective flux. The optimum composition for growth was found tobe La:Or, 2'0 g; SiOz, 0'6 g;

and NazWOr.2HzO,40 g. Cooling from 1150oC to 750oC was carried out at the rate of

loC/hr. The melt was then quenched from 750oC to room temperature and the remaining

flux dissolved by boiling in hot dilute NaoH solution. unreacted amorphous material was

894 JUN ITO

Tl.nr,n 2. UNn-Crr,r, Druewstoms on trrB Srr,rcera Aparrms (oxr)Synthesized at 1200'C in Air. (900'C for Lead Analogues and 1050.C for Mg

Analogues and Alkali Analogues)

Formula @ (A) , (A) I/ (A:; Formula o (A) , (A) y (A,)

Bazl,asSieONBazNdsSi6OmBazSmaSieO:sSrzl-asSioONSr:NdsSieO:oSrrSmsSioO:eSr:GdsSioO:oSrrDy8Si6OrSr:YsSieOueSr:ErsSioOze

Cazl-arSioOzrCazNdrSioOzsCa:SmgSioO:oCazGdsSioOzoCarDysSi6O%CazYsSieOzeCa2Er8Si6026Ca:LusSiaO:eCd:LaaSioO:ePb:LaaSLOzoPb2Nd8si6026PbzSmsSisOm

9 . 7 7 7 . 3 29 . 6 6 7 . 1 69.62 7 069 . 6 9 7 . 2 29 . 5 7 7 0 99 5 1 7 . 0 29 4 7 6 . 9 79 4 2 6 9 09 . 3 8 6 . 8 69 . 3 6 6 . 8 19 . 6 3 7 . 1 29 . s 2 7 . 0 09 .44 6 .939 3 9 6 . 8 79 . 3 7 6 . 8 19 . 3 4 6 . 7 79 . 3 3 6 . 7 59 .28 6 .689.64 7 .099 . 7 1 7 2 09 . 6 5 7 . r 29 s 8 7 . 0 6

605 Pb2cd8si6o26 9 .54 6 .95 548579 Pb2DysSi6O26 9 .47 6.87 534566 PbrYrSioOuo 9 .42 6.80 522586562 Mn2La8Si6Oz6 9 .63 7.08 568550 Mn:NdaSiaOzs 9 47 6 91 537541 Mn:>SmeSi6O:e 9.42 6.85 526530 Mn:GdeSisOze 9 .34 6.79 518523 Mn2Dy3Si6Ozo I .33 6 .7 | 506517 MnzYsSi6Ox 9 32 6.69 503571 MnrErsSioOzo 9 .30 6 . 65 498549 Mg2lasSisOm 9. 59 7 .05 561535 Mg2Nd8Si6O26 9 .45 6 .86 530525 MgzSmsSieOzo I .38 6 .80 518518 Mg:GdsSioOm 9.33 6.75 509511 Mg2DysSi6Or6 9.31 6 69 504509 MgzYsSioOzo 9 .31 6.64 498498 Mg2Er8Si6O26 9.28 6 58 491S7I NalasSieOm 9.69 7 .18 583588 NaYgSioOze 9 . 33 6. .7 5 509574 LiYeSi6O26 9 .34 6.72 508561 (Na,Th,La)16SioOm 9.66 7.13 576

separated by selective flotation. A mass of very fine transparent crystals (0.05 mm) wereobtained. Approximately 1.0 g (95lepure) was used for the chemical analyses.

Larger crystals of NazLasSioO:eF2 were grown as euhedral transparent crystals up to gmm in size with well developed faces using sodium fuoride flux by the combination of slowcooling (2oc/hr) an'd flux evaporation at temperatures from 1350oc to 950"c. The optimumcomposition for growth was found to be SiOz, 0.45 g; La2O3, 2.0 g; and NaF, 10 g. Shortprismatic crystals with {too} {tot } ana {oot } ate common. Also thick tabular crystalswith l11l I 1101 l ItOO ] and relatively large {001 } were formed.

The crystals of Nd (purple), Pr (blue), Sm (yeltow) and Gd (colorless) analogues werealso grown successfully. For non-alkaline species such as CazysSi6O26, potassium tungstateor fluoride were promising as flux.

Melting points of the compounds were not determined because most of the compoundsmelt incongruently and at temperatures higher than 1400oC.

LulrrNBscnNcB

Most of the compounds show a faint fluorescence in a variety of colorsunder ultraviolet excitation. rntense visible luminescense was observedin the yttrium and gadolinium analogues when activated by 5-10 per-cent of europium or terbium. The following compounds synthesized in


Terln 3. UNrr-Cnr-r, DrltrNsroNs op tnn PnospsATE- AND Borerr;-BoenrNoSrrrc.q:rn Aplurrs .q.Nl Htcrrr-v OxrorzBo Luo

Yrrnrulr Srr-rcerr Alenrns

Formula ao(At co(A) r (A') P,r

CasY2SirP+Oz(OH)zCaoYrSiaP:Or(OH):Ca:NarLaoSLPrOz(OH)zCa+NazlarSizPrOr (OH)z

Pb:a+Pbs2+P+SizOm

CasYzPoOroCaoY+SizPrOzoCanYoSinPzOroCaalazPoOzeCaoLa+SirPrOuaCa+LasSinPzO:eMgYrSisBOzoYroSirBzOzePbra+Pba2+YrSiaOroPbza+Pbr2+YrSisOzePb4+Pbf +Y6Si6Or6

CaYgSi5BOzo

s00'c 2 kb550'C 2 kb550'C 2 kb550'C 2 kb900'C 1 bar

1200"C I bar1200'C 1 bar1200"C 1 bar1200'C 1 bar1200'C 1 bar1200'C 1 bar1150"C 1 bar1150'C 1 bar900'C 1 bar900"C 1 ba,r900'C 1 bar

1150'C 1 bar

9 . 4 19 .399 . 6 29 . 5 19 . 7 89 . 3 79 . 3 69 . 3 5L489 . 5 79 . 6 29 . 1 89 . 1 59 . 7 79 . 7 29 . 5 99 . 2 8

6 . 8 76 .837 r r7 0 07 .336 .846 .846 . 8 26 .957 .027 .07o . / J

6 . 7 56 . 9 4o . 6 /

6 . 8 16 . 7 8

J L I

522570548607520519516541.)) /566491489. ) / J

J O Z

542506

air at 1280oC fluoresce: Ca2(Gd, Eu)3Si6O26, Srr(Y, Eu)sSieOro, and Na(Y,

Eu)gSioOro (scarlet red) and Ca2(Gd, Tb)ssi6or6 (bright yellow).

Visible emission spectra of the above compounds under mercury 3130 A

excitation are as follows.l Three europium activated analogues of

oxyapatite gave almost identical spectra regardless of their matrix ele-

ments. The strongest line was observed in the vicinity of 6190 A, and

fi.ve medium lines were detected at approximately 7050, 6500, 6260, 5980

and 5850 A. fne terbium activated analogue gave a completely different

visible spectrum; strong doublet lines were detected at 547 5 A and 5525 A,

and four medium lines were observed at 6300, 5880, 4950 and 4890 A.

ANarvsrs FoR HYDRoxYL IoNS

The presence of OH in CaaDy6Si6O24(OH), synthesized at 650oC, and 2

kilobars, was ascertained by infrared spectroscopy by Dr. Wilkins of

Harvard University. Its spectrum showed clear evidence of the vibra-

tion at 3540 cm-l while the anhydrous compound synthesized in air at

1050oC, presumably Ca2DysSioOuo, did not show this absorption' A later

study of Pb4Nd6Si6Orr(OH), and Pb42+Pbra+NdrSioOro by the same method

gave similar results.

1 Determination of the emission spectra was performed by Dr. L. Grabner at the

National Bureau of Standards.

JUN ITO

4 0 0 0 l o l o o 4 0 l o o @ € 1 o o 4 0 0 0 0 @o - d l N l 4 C ) + r l e l N - N N N N

' o 4 Q O Q I 4 S O O l r ) O O O O O O O. - i - C . C + r e - l < f F i 4 N N O F

4 4 4 o 4 Q Q O 4 l 4 O O o O l 4 O O 4 c ) 4 D D! n N + 1 I O I d r O O , @ L i i I O N i O N d N i

I

| @ 4 o o l o ' o Q Q O 4 n l c ) | o o 4 o 4 c ) O ol H N N O q D O O € l N N O N i i N e

- l 33R l g lBggg L99R- lR : -8RpR9

F > > t 5 5 U q c @ t r E> 7 7 = I a >i i i T i ' 7 i > > i

896

; i $ : N Q + 1 @ \ O < r N + : < ' c ) r \ O H € \ O + r

€qq4Q r i r€qqQ r{=qq tqqBEqqqq\ + r + r i . i ' r i o N N N N N N N e H H i * : i *

o

N

o N € N € r o , o 4 4 o \ N r r o n r o= O \ r - + \ O o - 4 N + . t N @ \ o N O N + oS r O c O + | t s ' O d r r i N - O O I O \ 6 o O € o o r r r. l

\ j + < r O O O O N N N N N N N i i i i i i i i

€ . _ r . * I S E € 8 9 3 R € $ $ S E R E F S BE \ q n 1 L a l ? n \ \ l " i l q q L q n q r \ \ \ \\ d ' < I O 9 O - I O N N N I T T N N N N

\ o 4 i o @ < r o hQ € T N O € 4 6

l q a n n n \ \ 9. ; + q N - o r + €^ r € € 6 N € - 4 N . O o' : N o @ € o € r r N N

" c + < r o o e a N N N N N

' : € o € F E S S S 3 3 S S 3 Y - o o n 4 o i3 iq i \ n t { i t i i : i 4qqs{qEq\ < 1 + O O A N N N N N N N i i i i i i i i

u i - - - \ O { = - c O o o N o O o . r ) + O \ D \ O * s5 O O ' O N o O " 9 & r O N @ O \ + e ! 6 a N 6 @E - : q a I = : n n \ a . . : q q q o . * * n { \

\ < ' < ' O C " ) O N N N N N N N H

N r . N n 3 3 E S H S H pl a o l ' : " ? q - . l n

N N N N N E i : F F F *

$ €s r

A Y" . O d t i i

%B F q'\ -:: .YS + i N N

Z Y

5 QA 2 - . ^v- i6 i . ;

o?^ ".. n

: i J i N6 A

O Y

+ @

A P

V N

a ( \

I

" i ' ^

b o h =. A ? , >

! 9 0 9

€ N O \ D € \ O < 14 0 0 \ o o : o ,o \ a 6 6 6 € r

< 1 \ t \ o o € N 4 \ o@ O @ i c " ) i < . r Oi o \ c a @ l o N i H

. l€ N N N N N N N

H N N

g * - -' 6 l r q a

\ d r d l o

r :O , r

z

m

a

,-i ,

E OE " rp <v z

< 3X

z 4< H

z a .! <

-4 z

< r'lzxz z< H

t1 -f l I [

z ^a 3a A

3 . <s z

A

d.

X

Fl

a

Y oO. .O

l sF

\ O N

o . , r

N €

C \ N

{ d

O O o . c O + + € € O \ ' O O@ i a n 4 0 N : @ @ r

' o d r d r o o e o o o N N N

- O O - N N O - N O+ F c ) F . = 9 F - - c. q - N i ( t N N - . / )

= 3 : 8 R : R : F = S =a - - + o i o . o 6 i r j . n + 5


Tnsrr 5. Rosur,rs oF H2O ANALYsEs

897

FormulaHzO analyzed temperature range

100-180'c 230-650"C 1050'c

TheoreticalHtO

a SrrEroSioOzr(OH)zb CarLueSioOz(OH)zc. SrnYeSioOn(OH)zd. Ba+GdoSioOz(OH)ze MnaSm6Si6Or(OH),f. Pb4Nd6si6or4(oH)'g. Pb2a+Pba2+Nd4Si6O26

o 23ok0 o3To

0 . 7 %0 . 8 %1 . r %| 27oo . 7 %

0 . e %1.0'ro1 1 %o 87okt . 0 %0 8 %o.0,%

0 6s%0.o4Ta

Analyses a, b, c, d, and e were done by Mr. Ralph Paulson at the National Bureau

of Standards using a carbon-hydrogen analyzer at 1050'C. Analyses g and f were carriedout by Mr. W. Sabine at Harvard Universitl., a micro-coulometric technique.

Results of the analyses for HzO are given in Table 5. All the hydroxyl-apatites heated at 1100oC gave X-ray powder patterns indistinguishablefrom those of the oxyapatite Mrz+1,1ra+516O26. In the cases of lead ormanganese analogues, formation of oxyapatite could be accompanied byoxidation of the cations to compensate for the loss of hydrogen; in othercases where oxidation was not possible, valency compensation for theloss of hydrogen was achieved by changes of cation proportion from theoriginal hydroxylapatites.

Differential thermal analysesl were also made on Pb4Nd6Si6O2n(OH),and Pba+zPb2+aNdrSioOro. The former showed an endothermic peak at350'C while the latter did not show any sign of dehydration. X-raydiffraction powder pattern taken of the former sample after the experi-ment showed the formation of oxyapatite indicating that the endothermicpeak was due to the dehydration of hydroxylapatite.

RBsurrs AND DrscussroN

Compositional range, nomenclature and, solid solubility. A large variety ofsilicate and phosphate compounds having the apatite structure, includ-ing some containing borate, were prepared (Tables I,2, and 3). Divalentions from Mg'+(0.66 A; to nu,*1t.34 A) can replace Ca2+. Monovalentions such as Li+ and Na+ can also replace Ca2+, but K+ can replace Ca2+only partially. Valency compensation is achieved by introduction ofrare earth,lanthanum, yttrium or thorium ions.

I Difierential thermal analvsis was oerformed bv Mr. E. M. Levin at the National

Bureau of Standards.

898 ]UN ITO

Over sixty chemically different compounds which gave single phaseapatitic X-ray patterns have been synthesized. The unit-cell dimensions,together with the conditions of the syntheses, are given in Tables 1,2,and 3. Unit-cell dimensions a and c of hydroxyl analogues are plotted interms of ionic radii of divalent cations and rare-earth ions in Figures 1 and2 respectively. The natural minerals can be located in their approximatepositions in these diagrams according to the chemical compositions andtheir unit-cell dimensions. Of the three natural minerals lessingite is theclosest to the end composition CaqLaoSioOzr(OH)r. Britholite and becke-lite resemble lessingite but contain small proportions of the isomorphouscompounds Na(Ln,Y)eSi6Or4(OH), and Caro(POn)u(OH)r. Abukumaliteis close to the end composition Ca+YoSioOz+(OH)2.

Complete solid solubil ity exists between hydroxylapatite Car6(PO)6-(OH)r, and CaaYoSi6O24(OH), (abukumalite), under hydrothermal con-ditions (Fig. 5). Solid solution is incomplete between Caro(PODo(OH)zand Ca+LaoSioOz(OH)z (lessingite) under the conditions I used, owing tolarge differences in their unit-cell dimensions. A complete series be-tween Shs(POr6(OH), and SrrLaeSi6O24(OH)2 is probable because of thesimilar sizes of Sr and La.

The following solid solutions of oxyapatite based on 26 oxygens percell also have been investigated: CazYaSioOzodCasYzPoOzo, Ca2La3Si6O26PCasl-azPoOzo, CazYsSioOredYroSirBzOzo and MgzYaSi6O262ytoSioBrOx.

C e

. N do s .

Gd

" D Y' E r

L u

l . l 4

1 0 7

l o 4

r 0 0

9 7

9 2

. 8 9

8 5

o , r E s s t N c t r E 9 . 6 7: j O l t c x E l l r E 9 6 6

a R t t H o ! l T I 9 _ 6 1

S rC oM n

I O N r C R A D I I i ( l h r " n r )

Frc. 1. Chemical stability region of hydroxy-silicate-apatites: Mr2+(Ln)6r+$iu6z(OH)tin terms of ionic radii (Ahrens, 1952) ol the divalent cations and lanthanides. Unit-celldimensions o are given as contours. Natural minerals are plotted according to the approx-imate chemical compositions. Pb analogues given in open circles do not fit the contours.


t l 4

t o 7

1 0 4

l o o

9 7

9 2

8 9

8 5

99 1 .12 1 .20 1 .34

M n C o S r P b B o

t O N r C R A D I I i ( A h r " n r )

Frc. 2 Chemical stability region of hydroxy-silicate-apatites: Ma'?+(Ln)s3+SisOzr(OH)z

in terms of ionic radii (Ahrens 1952) of the divalent cations and lanthanides. Unit-cell

dimensions c are given as contours Natural minerals are plotted according to the approxi-

mate chemical compositions. Pb analogues given in open circles do not fit the contours.

The unit-cell dimensions have been plotted versus chemical compositionsand are given in Figures 6, 7, and 8.

Oxyapati.te and, hyd,roxylapatite. Unit-cell dimensions a and c of oxyapatiteare given in Table 2 and unit-cell volumes are plotted in Figure 3 in

terms of ionic radii of divalent ions and rare-earth ions. The chemical

I . l 4 L q

1 o 7 C e

o 1 0 4 N dA r o o S m

e 7 G d

tr rr,

8 9 E r

' 8 5 L u

l . t 4

t.o,a

t.o o. 9 7

. 9 2

. 9 0

. 8 9

8 5

/ng iAn Co Sr Pb Bo

I O N I C R A D I I b y A h r e n s i

Frc. 3. Chemical stability region of oxy-silicate-apatites: M22+(Ln)g3+Si6O26 in terms of

ionic radii (Ahrens 1952) of the divalent cations and lanthanides. Unit-cell volumes are

given as contours. Pb analogues in open circles did not fit contours,

L o

NdA .

J M

Gd

'31Lu

900

. 8 5 . 8 9 9 2

[ u E r Y D y

JUN ITO

. 9 7 t . 0 0 I 0 4 1 . O 7

G d S m N d C e

V

A3

l t 4

I o

I O N T C R A D I I A

Fig. 4. Unit-cell volumes of Ca analogues of hydroxylapatites CarLnrSioO:n(OH)z andoxyapatites CazlnsSi6ozo versus ionic radii (Ahrens 1952) of lanthanides.

6.90

6.85

6.8 0

9.15

9.10

9.3 5

CoaY5 SioO2a(0t l l2 Co16P602 a (OH12x

o

c o t o - , Y , P 6 - , s i r o 2 4 ( o H ) 2

Frc. 5. Unit-cell dimensions of solid solution series of apatite CaropoOzr(OH)r-Ca4y6si6-Oz(OH)z abukumalite.

G

A

oA

c,i b'

coY

SILICATE APATITES AND 1XVAPITfiBS

c.cd P602.(oH)2

CorY.Si.O25 CorY2 PaOra

g ? , ! 9

c o r . t Y r - t s i a - t P t o z '

Frc. 6. Unit-cell dimensions of solid solution series of Ca2Y6Si6O5-CasYzP6Ozs.

stability region of oxyapatite is considerably greater than that ofhydroxylapatites (Figs. 1 and 2).

Unit-cell volumes of Ca analogues of hydroxylapatites and oxyapatitesare also plotted against ionic radii of rare earths in Figure 4, showing thatthe series of hydroxylapatites and oxyapatites are different. Distinctionbetween the hydroxyl and oxyapatites was also clearly indicated by thechemical analysis (Table 5) and infrared spectroscopy discussed in pre-vious sections. Cockbain and Smith (1967) did not observe these dif-ferences in their synthetic compounds, which probably are anhydrous.

The compounds synthesizedby Trcjmel and Eitel (1957) are also prob-ably oxyapatite because the present study has demonstrated that it isunnecessary to postulate retention of water in the silicate compoundssynthesized in air either by solid-state reaction or by flame fusion athigh temperatures. These authors state that the role of water is uncer-tain because of the difficulty in analyzing for small amounts of water.X-ray data obtained by Trcimel and Eitel are not identical with thoseobtained from the present study or those given by Cockbain and Smith(1967). Some differences are greater than can be accounted for on thebasis of errors in measurement.

901

c

A 6 . 8 5

Co,fiOr.(oH)z

o 9.40

A9.35

9 .30

902 TUN ITO

Cor l c rS i 5Oro9 ?

C o r l o r P o O 2 6a 6

c

A

6.90 Co,oPoOrJOH l,

Co,f.Or.(OH),

9.70

o

A

Co2* , Lo6 _ , S i 6_ ,px Oz e

Frc. 7. Unit-cell dimensions of solid solution series of CazlasSiaO:s-CaslazPoOzo.

Chemical analysis of sodium Ianthanum silicate oxyapatite, grown bythe tungstate flux method, gave LazOz 78 percent, SiO, 21 percent andNazO 1.6 percent (analysis by Rains and Ito). The formula derived fromthis analysis Nao.gaI,ae 65Si6 aO26 is close to NaLagSieO26, suggesting a com-pletely oxygen filled structure. In each synthesis of the same cation com-bination, the resulting oxyapatites, except lead analogues, gave exactlythe same unit-cell dimensions, even though the cation ratio in the start-ing mixtures varied. For instance, mixtures with empirical compositionsCazYaSioOzr(O)z and CaaY6Si6O2(O) after heating at 1250oC, gave iden-tical unit-cell dimensions and X-ray diffraction intensities. The samefindings were reported by Cockbain and Smith (1967) with Ca-Laanalogues. Their runs containing Ca and La with the ratio of 4:6,5:5and 6:4 gave nearly identical unit cells. I can not agree with their ex-planation: "the lattice parameters of Caal-a6(SiO+)oOH and Carl-ar(SiOa)oare virtually identical, since the substitution of La3+ for part of the Ca2+has expanded the lattice sufficiently for (OH)- or F- oravacancy to be ac-commodated without strain." We have clearly demonstrated (Fig. a)the differences between hydroxyl- and oxy-compounds. The above ob-servations suggest that there is only one oxyapatite (Table 1 and 2, Fig.


co2Ygsi6o26 vrosl-Ugzo2 t . 5 t o . 5 o

Y16_q!o, S i4*xB2_xO2o

Frc. 8. Unit-cell dimensions of solid solution series of CazYsSieOrs-YroSLBzO:0.

4) for each combination of cations, in which the loss of hydroxyl ion is

compensated by cation replacement, and that stabilized oxyapatite con-tains 26 oxygens in each unit cell, as originally postulated by Korber andTrtjmel (1932) and McConnell (1938).

It has Iately been known that the simple rare-earth silicate, LnaSiaOrz(Ln:La, Pr, Nd, Ce, and Sm) (Keler, Godina and Savchenko, 1961;

Leonov, 1962; Miller and Rase, 1964), has the apatite structure. Kuzminand Belov (1965) reported that this apatite is cation-deficient oxyapatiteand suggested the formula Lag.szSioO:6. This interpretation fi.ts well withthe results of the present experiments because unit cell dimensions of theabove compound o:9.72; c:7.18 are significantly larger than those ofNalagSioOzo a:9.69; c :7.17 . I t seems highly unl ike ly that the uni t ce l lof LasSioOza is larger than that of NalagSioOzo. The finding of cation-de-fect apatite opens the wide possibilities of cation deficiency in silicateapatites generally. Further investigation is indicated.

903

904 JUN ITO

9.8 0

9.7 0

9 . 6 0

9.5 0

9 .1 0

6 . 9 0

6 . 8 0

tttlr,fltrsoo.o , , rfiri'oo.o

O t . X 2 g

rufi, ru!.-, yz, zxs i 6 02 6Frc. 9. Unit-cell dimensions of solid solution series Pboz+pba4+y2si6or6-pbl+yeSi6ore.

Pure calcium phosphate oxyapati te Ca16P6O26 on my interpretat ionshould not exist because the26 oxygen sites must be f i l led for the apati testructure. fndeed infrared examination showed a strong OH absorptionpersist ing in a synthetic apati te heated in air at 1100"C for 48 hours. I foxygen deficient Ca oxyapati te exists, i t was not found as an end memberof the sol id solut ion series extended from the oxyapati tes Ca:LneSioOzo toCasLn2P6O26 (Fig. 6 and 7) . However, should oxygen-deficient Caoxyapati te be stabi l ized by some mechanism, the extrapolated values ofthe unit-cel l dimensions as given in Figure 6 and 7 lead to predictedvalues of cel l dimensions of oxygen deficient Ca-oxyapati te of a:9.38+0 .014 , c : 6 .85+0 .01A .

Unit-cell dimensions of lead analogues of oxyapatite synthesized inair vary depending on the rare earth to lead ratio and on the tempera-tures of the synthesis. Deficiency of oxygen was at first suspected, butthat hypothesis was later discarded. The experimental data fit the exis-

o

A

c

A


tence of a continuous series from Pbz2+YeSioOro to Pbr,2+Pbaa+YzSioOm

(Fig. 9). Valency compensation is achieved in part by oxidation of diva-

Ient lead ions to higher valency. It has been found that the highly oxidized

compound is converted to the less oxidized compounds at elevated tem-

peratures due to evaporation or formation of a lead silicate accompanied

by partial reduction of Iead to the divalent state. The compound Pb22+-

YaSisOro, containing only d.ivalent lead is stable up to 1150'C and de-

composes to yttrium sil icate and vapor at 1200'c. compounds contain-

ing more lead, including oxidized lead, shrink in volume on heating at

temperatures over 1000oC, but melt before enough sublimation has

taken place to reach the stabilized end composition of Pbz2+(Ln,Y)sSioOzo.

Presence of higher valency lead was tested qualitatively by dissolving

fi.nely ground samples in 3N acetic acid solution containing potassium

iodide. Iodine liberated by the reduction of Pba+ gave intense orange-

yellow coloration to the solutions.Oxypyromorphite described by Wondratschek (1963) as PbroPoOzs and

lead oxysilicophosphate-apatite by Dietzel and Paetsch (1956) of sup-

posed formula PbroSizP+O%, were also resynthesized and identified as the

oxidized compound containing higher valency lead ions. The formulas

may be rewritten as Pb82+Pb24+SizPaOzo and Pbg2+Pba+P6O26, respec-

tively.The writer suspects that manganvoelckerite, a divalent manganese

bearing oxyapatite (Wickman, 1954; Mason, l94l), also may not be an

oxygen deficient apatite. Its dark reddish brown color and its relative

instability indicate the presence of higher valency manganese ions.

From the above results, the writer proposes the following general

formula, (M+M2+M3+M4+)rg(Si4+Ps+$e+)6026 for well stabilized oxyapa-

tite in which both cation and anion sites are filled. The interpretation of

compounds studied in this paper does not require the assumption either

of oxygen deficiency or cation deficiency, and these substances can all be

explained by overall charge balance. Nevertheless, the possibility that

oxygen-deficient oxyapatite might exist has not been excluded, par-

ticularly under extremely unusual conditions or compositions. Cation

deficiency can also not be ruled out, in fact, cation deficiency is more

likely to be found than oxygen deficiency.

Steenstrupine. Chemical composition of steenstrupine (Strunz, 1942;

Machatschki, 1943) a complex alkali, alkaline-earth silicophosphate in-

dicates this mineral should be found within an isomorphous series from

Caro(POe)o(OH)z?Canln6Si6O%(OH)rPNazl-nsSioOr(OH)2. No distinc-

tive phase was found at the steenstrupine composition. Only a mixture of

the silicate apatite and monazite phases were observed under diverse

906 JUN ITO

conditions tried by the writer. Published X-ray powder data by Sabinaand Trail l (1960) of the specimen from Kangerdluarsuk, Greenland, wereidentif ied by the writer as a mixture of apatite and monazite. The hy-drothermally recrystall ized steenstrupine specimen from the same lo-cality preserved at the Mineralogical Museum at Harvard University alsogave the X-ray pattern of a mixture of apatite and monazite.

AcKNowLEDGMENT

This work has been carried out partly in the Department of Geological Sciences, Har-vard University and paltly in the Inorganic Materials Division, National Bureau of Stan-dards. Thanks are due to the U.S. Advanced Research Project Agency for grants whichmade this work oossible.

The writer is indebted to Mr. H. s. Peiser, for his help in revising the original manuscriptof this paper and to Professor C. Frondel of Harvard University for a critical reading of themanuscript,

The writer also expresses his appreciation to Dr. R. Roth and f)r. T. Negas of the Na-tional Bureau of Standards for valuable suggestions. Also, the writer wishes to thank Dr. AKato at the National Museum of Japan for supplying an abukumalite sample and X-ra1,data.

RrlonrNcrs

AHRENS, L. H. (1952) The use of ionization potentials, r. ronic radii of the elements Gea-ckim Cosmochim. Acta,2, 155-169.

cocrnerN, A. G. ANn G. v. Surrrr (1967) Alkaline-earth rare-earth silicate and germanateapatites. Mineral. M ag, 36, 411-421.

Drerznt, A. axn H. Pnorscu (1956) Untersuchungen tiber das System pbO-SiOrprOs

Glastech. Ber., 29, 350 355.Fnowont, C. (1961) Two yttrium minerals: spencite and rowlandite. Can. Mineral.,6,

576_581.Ge'v, P (1957) An X-ray investigation of some rare-earth silicates: cerite, lessingite,

britholite, beckelite and stilwellite. MineraL. Mag ,31,455 468.H,tra, S. (1938) Abukumalite, a new mineral from pegmatites of lisaka, Fukushima pre-

fecture. Scl Pap. Inst. Phys. Chem Res., Tokyo,34, 1018-1023.HAcrr.n, G eNo F. M.q,cnarscnrr (1939) Der Britholith ist ein cereden-sitikatpatitl

Zentratrbl. Mineral , A", 165-167.

Jernr, H. w' lNo v. J MorrNsxr (1962) spencite, the yttrium analogue of tritomite fromSussex county, New Jersey, Amer Mineral , 47,9-25.

Knrnn, E. K., N. A. Goorxa aNn E. P. Sevcnn'xo (1961) Reaction in the solid phase ofsilica with oxides of rare-earth elements (LazOr, NdrOs, Gd:Os) 1zo. Akad.. NauhSSSR, Ser. Chem. lO,1728-1735.

Konren, F. axn G. 'rnolrnr. (1932) untersuchungen iiber Kalk phosphosaure und Kalk-Phosphosaure-Kieselsaure-Verbindungen. Z. Elektrrochem., 38, 578-582

KuenrveNova, r. r. aNn G A. Sroonnxro (1963) Minerals of the britholite group. Dohl.Akail Nau.k,.S"S.9R, 148, 912 915.

KuzltrN, E. A. aNr N. v. Bnr,ov (1965) crystalline structure of the simplest silicates of Laand Srn. Dohl. Aha.d. Nauk. SSSR, Sey. Math. P/zys. 165, 88-90.

LroNov, A. r. (1962) The valence of cerium in synthetic and natural cerium aluminatesand silicates, Pt. 2, cerium silicates. Izr. Akad. 1[oz& s^ssR, ser. chem. lz.2o}4-zogg.


Mecn.lrscurr, F. (1939) Sind Abukumalite und Britholith Glieder der Apatitereihe?

Z entr aJbI. M iner al, 1939A, 16l-164.-- (1943) Steenstrupine ist kein Silikat von Formeltypus Apatit. Naturwi.ssenchaJten,

31,438-439.M.rsoN, B. (1941) Mangualdite is manganvoelckerite. Geol Foren. Forh.,63(4), 383-386.

McCoNNnLr-, D (1938) A structural investigation of the isomorphism oI the apatite

group. Amer. Mineral.,23, l-t9.-

" (1965) Crystal chemistry of hydroxyapatite, its relation to bone mineral. Arch oral'.

Biol , lO, 431-+36.

MTLLER, R. O. eNn D. E. Rasa (1964) Phase equilibrium in the system Ndro3-sior. "r.

Amer. Ceram. Soc , 47,653-654

Monoznwrcz, J. (1905) Uber Beckelith, ein Cer Lanthano-Didymo Silikat von Calcium

Tschumaks Mineral. Petrogr. Milt., ?4, l2O-134.

Nrcrranva, E. A. lto I. D' BnoNBu.q,N-STARYNKEVICTT (1956) Britholite in Skarn rocks of

Western Transbaikalia' Z ap. M iner al,' Ob shch., 85' 509-514'

NnuuAN, H., T. SvotonuP AN; P. C. SeBeo (1957) X-ray porvder patterns for mineral

identification. III. Silicates. Norshe Vidensh. Akad,. O:lo, 1., Mat.-Naturtt. Kl'. p.6, 17 '

plates 19 and 30.

ouou, K. ero S. Heseclwl (1953) Yttrialite and abukumalite from pegmatite of Suisho-

yama, Iisaka village, Fukushima, Japan. J . J ap' A ss' M inerol P etrol'og1- Econ' Geol"'

37,2l -29.Rocrns, A. F. (1912) Dahllite (pedoiite) from Tonopah, Nevada, voelckerite a new basic

calcium phosphate. Amer. J. Scl., 33, 47 5-482.

Ruownve, A. V., R. NrrrrrN AND N. V. Brlov (1962) Ceftosil, a cerium bitholiLe Dokl'.

Ahad. Nauk SS.SR, 146, 1182-1183.

Sarrwa.,A.P. ,woR.J.Tnarr , r , (1960)CatalogueofX-raydi f f ract ionpatternsandspecl-men mounts on file at the Geological Survey of Canada' Geotr' Surtt Can' Pap' 6O-4'

98.

Sa runa r .K . ' n .NoA .Ka ro (1962 )Abukuma l i t e f r omSh inden ,G i f u , JapanJ .M i . nc ra l .Soc . JaP . , s ' 328 334 .

srnuNz, H. (1942) Steenstrupine, ein silikat von Formeltypus Lpatit' Noturwissenschaften

30. 65.

Tn6urr-, G. (1932) Untersuchungen iiber die Bildung eines Halogen-freien Apatits aus

basischen Calciumphosph aten. Z. P hy sik, C h em', 158, 422-432'

Tnrilror,, G. aNo W. Errnr. (1957) Die Synthese von silikatapatiten der Britholith-Abuk-

umalit Gruppe. Z. Kristallogr.,lO9,231 239'

WrcrulN, F. E. (1954) Minerals of the Varutriisk pegmatite, XXXVII, Manganvoelcke-

rile. Geol. Fdren. Stockhol,m Fdrh,76, 495-500'

wrNrunn, c. (1901) Britholith, ein neues Mineral. z. Kri,stallogr. Mineral.,34, 685-687.

woNonlrscnex, H. (1953) Untersuchungen zur Kristallchemie der Blei-Apatite (Pyro-

Manuseribt receizted', Nottember 3, 1967; accefted, for publicatioto, February 8' 1968'

the american mineralogist, vol. 53, may-june, 1968 · 2007. 4. 4. · the american mineralogist,...

Documents