Gems Tourmaline

8/3/2019 Gems Tourmaline

1/10

SPECTROGRAPHIC ANALYSIS OF TOURMALINES FROMTHE ISLAND OF ELBA WITH CORRELATION OFCOLOR AND COMPOSITION

Gurpo Canonnr aNn RnNzo PrnnuccrNr, Institute of Mineralogy and,Geochemistry, Uniaersity oJ Florence,Florence, Italy,

INrnorucrroNThe colors of the beautiful crystals of tourmaline from the Island ofEIba and their zonal coloration have always attracted the attention ofnaturalists. Dolomieu (1) in 1798 examined the tourmalines of Elba de-scribed as "half white and half dark";Ottaviano Targioni Tozzetti (2)(the son of Giovanni Targioni Tozzetti, the great Tuscan naturalist of1700)described or the first time the tourmalines of different colors thatLieutenant Ammannati collected when he was garrisoned on the Islandof Elba.Antonio D'Archiardi (3), in his "Mineralogia della Toscana," recordedsome nteresting data on the succession f colors on the crystals and onthe variation of density with color; Giovanni D'Achiardi (4) wrote his

doctor's thesis on the tourmalines from Elba, after examining about 6000crystals or fragments of crystals.More recently, E. Grill (5) found some new forms on the tourmalinesfrom the Island of Elba. From the crystallographic studiesof G. D'Archi-ardi, we note a regular variation of the axialratio with the color; we have,for example, the lowest value of the axial ratio for the yellow-green our-malines (a:c:0.441154), then follow in order the dark, the colorless,the pale pink, the dark yellow, the greenishyellow, and, lastly, the yel-low tourmalineswith a:c:0.454079. Consequently,he color varieswiththe chemical composition of the crystals, as is also shown by the specificgravity and by the refractive indices. The black crystals have the max-imum density (3.t67-3.174), hen come the greenishyellow, the palepink, the colorless (3.017), and the yellow, which are the lighest of all(2.e50-3.014).It thus appears hat the succession f colors s very similar to that ob-tained by arranging the crystals according to the increasing value of theratio a,ic.When we consider the double refraction, we can show that the color-Iess crystals must be classified with the slightly pink and the greenishcrystals. We obtain the following succession: ink (co-e :0.0200 for yel-low light), colorless(0.0204),greenishyellow (0.0220).We may conclude hat the color, or at least some colors,are influencedby the chromophoreswhich enter in solid solution to a considerableex-

tzt


2/10

I22 GUIDOCAROBBIAND KENZO PIERUCCINItent. For some colors (black, green,greenish yellow, and red), the colorcannot be attributed to minor constituents present in such small quan-tities that they could have no influenceon the axial ratio and other physi-cal characters.

The structure of the solid solution forming the tourmalines has beendetermined by M. J. Buerger and W. Parrish (6) on crystals from theEtta Mine, South Dakota, to which they have assigned he spacegroupC$,.Miss Kulazzewski (?) and Machatschki (S) had previously reportedon the structure of tourmaline. The formula of the solid solution mav bewritten, according to Buerger and Parrish, as:(Na, Ca)R:(Al, Fe)6B3SioOzr(O,H, F)4

and in a general way, according to Machatschki:XYeSj6Ba(O, H, F)31, X:Na, K, Ca, Mn) , (Y:Al , Fe)

The elementsmust be grouped together, as in all minerals, on the ba-sisof ionic radii. Sodium may be replacedby calcium and manganese;9Yis never represented n nature by 9Al, but instead by Alsl-i or Al5(Mg,Fe)a. The lithium is almost always present, and from the point of viewof isomorphism there is equality in the group of the two atoms LiAl with2 Ms.

For the Elba tourmalines, there was lacking a spectrographic researchof the minor constituents, which was necessary or a study of the rela-tionship of composition and color.Spncrnocn.q.Purc NALYSTs

We have prepared spectrogramsof tweniy differently colored samplesfrom the black, green and red, to the greenish yellow, all coming fromthe Museum of this Institute. Every one had been separated from aclear and transparent crystal and carefully examined under the micro-scope; the very fine powder was crushed in clean mortars, dried in anoven, and then vaporized in the arc of very pure carbon electrodes0.6centimeters in diameter of the Ruhstradt firm. The substance (40 mg.)was placed in a cavity .2 cm. in diameter and t.2 cm. deep, and volatil-ized in the arc with 190V and 8A Ior 6'45".We have useda ZeissQu 24 spectrograph;the field of observation wasbetween 5000 and 2300 A. The focusing adjustment of the slit was for2300A. We used Hartmann's diaphragm T 10 with nine windows and asintermediate screen the D5. The breadth of the slit was 0.006 mm. Weused Ferrania Cappelli ultracontrast orthoplates and developed for fiveminutes with methyl hydroquinone in the developer.


3/10


4/10

124 GUIDO CAROBBI AND RENZO PIERUCCINI

i r eI

o l I^ l6 l- i

l - r l kR I i l xx#x i t -F>FE#x>Fi tE f iEExx i xg t . iEEE#x -EF r E iEE

l * * -x i t> io#F>x :F i r l i lII xxn* - | -g> f i i , l x -EnE I ti1 >4q i4,9t . "l x x o x >

F

Fr | r F r >

F

-E

XFi

Er

a lFr

F

x x f x# t -EEEE-x>E l rE i i I hFr

F

ll

ila

I

E

a

tlXE

u

l r i| . : I X X t T X Ft - l5 . - ,l - F E l i F x > i i I l F # |

r# iH t igx=F l r r iE rl a F > I F T F X o F F I I F F '

5 . 3 !l o i E t i # x > F F t t . F tl - f i > l F - x E l f , i l - g II - F E f i F # x H F F I I F d Il - i o F F . x E F i i

* xax - | -EoF I : x> I l tx x > x 9 l - F s F i - x a i i

5 . 5x x E x 9 l - F o i F E x - i i Il kl x x > x E l o r : o f i # x # F i iIx x > x o l o f i o E i - x g l i E t . # t

l x xExo l o F E F 9 - x # l f i F l - g Ix x E x o l o i > E F , x # l F i l - i E

il

Fr X X t r X >X X > X >

" : 3- &t -- | E al . ; *

I I F F F

6 2 ,

X X E X O E O i E E S O X # I F # I q > F j' ca ' i ? ' t gd ES g v E; i s g > d I "E


5/10

SPECTROGRAPEIC ANALYSIS OF TOARMALINES FROM ELBA 125GBocnnurcer Corqsrpnnarroxs

Al l the minor constituents determined by spectrographic means maybe consideredas isomorphoussubstituents of the major constituents ofthe tourmalines. Their presence s thoroughly justified by the followinggeochemicalconsiderations.It is very well known that lithium, owing to its too small radius (0.78),does not replace the sodium or potassium in minerals, but substitutesfor magnesium (0.78). It is found, therefore, n small quantities in dio-rites, gabbros,and in more basic rocks (9); in the acid rocks it forms itsown minerals, such as spodumeneor amblygonite, and it partly substi-tutes for magnesiumas in tourmalines (Al Li : 2Mg).While rubidium substitutes for potassium and is enriched in the feld-spars, particularly in amazonite (up to 2.74Td, caesium yields its ownminerals in the residuum of the crystallization of the acid rocks, or ispresent in traces in the potassium minerals, for it has a radius very dif-ferent from potassium (K:1.33, Cs:1.65). Rubidium is present n themicas to a larger extent than is the caesium (muscovite 0.4OToRbzO;0.OlXToCszO) 10). Therefore,caesium concentrates tself more than therubidium in the residuum of the crystallization of the granite of theIsland of Elba, and we find it in the tourmalines, while rubidium isabsent.

Barium found by us in every sample is a replacement of potassium(Ba 1.43, K 1.33);i t is known that it is contained n orthoclase up toI.5%).The maximum quantity of barium is therefore in the syenitesand trachytes (0.18le) (11); the nepheline syenites and the phonoliteshave only 0.05870, the granites and the liparites 0.048To, the dioritesand andesites O.0267o,and the gabbros and the basalts 0.007/6. Thebasic rocks deprived of potassium are also deprived of barium. In thediorites, barium is contained in the biotite (up to 3/); the calcareousrocks have barium only in small quantities.Beryllium, owing to its small radius (0.34), is a minor constituent ofacid rocks where t substitutes for silicon (0.39);the quantity of berylliumis greatest in the pegmatites (12), where it forms its own minerals; itsisomorphism with magnesium s-limited, and so in the basic rocks, it isabsent or present in quantities smaller than 0.00I/s. fn the crystals oftourmaline that we have studied, beryllium substitutes for silicon and isalways present.As in the first two groupsof the periodic system, a geochemicalanalogydoes not exist between sodium and potassium, and between magnesiumand calcium, so there is no analogy in the third groups between alum-inum and scandium. Scandium is a substitute for magnesium, and it ishidden in the iron and magnesium minerals (13); al l the magmatic


6/10

126 GUIDO CAROBBI AND RENZO PIERUCCINIolivines contain scandium, as do augite, diallage, and other pyroxenesand amphiboles Mg 0.78 A, Sc 0.89,Zr O.87).*Therefore, n the tour-malines, the quantity of scandium present s larger in the green ron andmagnesium-rich crystals. Titanium is present in all analyzed samples,and it is a substitute for iron and silicon. The geochemistry of niobiumand tantalum is not very well known, but tantalum should have a be-haviour like phosphorus and in the tourmalines substitute for silicon.Copper is present in the tourmalines as an isomorphous substitute formagnesium.

In the first group of the periodic system, Cu, Ag, and Au are clearlychalcophil; Iikewise in the second group, Zt, Cd, and Hg; and in thethird, Ga, In, Tl; but the first element of the three series s often hiddenin the minerals of the rocks. Copper in very minute traces is very com-mon as a substitute for the isomorphous group Mg-Fe" in the silicatesof the rocks. This substitution has not been studied sufficiently, andcases of isomorphous substitution of magnesium by copper in simplecompounds are not known in chemistry.Many years ago, one of us (14) demonstrated that in the very wellknown double nitrate of the series:

2X(NOJB'3Mg(NOg)z '24I j2O,X:Ce, La, Nd, Pr , Sm) ,magnesium can be partially or entirely replaced by copper and also bycadmium.

MrNon CoNsrrruBNTSAND CoronIn the solid solution which constitutes the crysta ls of tourmaline thatwe have studied, the presenceof the minor constituents that we havefound is explained by their isomorphism with some major constituentsl

there remains to be determined the relationship between minor con-stituents and color.The question of the color of allochromatic minerals is one of thepartially solved mineralogical problems. Formerly, we always attributedthe color to the presence n the crystals of elements which usually gavecolored compounds; only recently have we noticed in colored allochro-matic minerals the absence f coloredcompounds,and the color was thenattributed to a physical phenomenon,as ionic deformation in the crystal.Kraatz, Koschlov, and Wcihler (15) thought color was due to inor-ganic pigments. Weinschenk (16) found titanium in the smoky qtartz,and he prepared some colored corundum with inorganic pigments.* Newhouse has also reported tlat magnetite of granitic pegmatites contains scandium

and not the others. (Carnegie nst.Wash. Publications, Year Book, No.40 for the yearL94041, pages 142 to 144. Issued December 12, 1941.)


7/10

SPECTROGRAPHIC ANALVSIS OF TOURMALINES FROM ELBA r27Brauns (17) was the fi.rst to use the absorption spectrum for theidentification of the chromophores.Then followed Weigel and Habich(18), and especially Kolbe (19), who compared the absorption curve ofa number of chemical compounds n pure crystals and in solutions withthose of various allochromatic minerals.In particular, he studied the absorption of the compounds of Mn, Cr,Fe, and Ti. From such comparisons, t appearsprobable that titaniumwould produce part of the color of amethystl ferrous iron, of the greentourmalines, of clinochloreand of xanthophyllitel ferrous and ferric iron,of blue sapphire, spinel, and of green corundum; chromium, of the red

and green spinel, of alexandrite, blue disthene and of the emerald; tri-valent manganese,of the pink tourmalines and of red almandine. Theabsorption curve of the pink tourmalines presentsalso a certain affinitywith that of potassium permanganate.But the question of the color of allochromatic minerals is certainlynot concludedwith the investigations by meansof the absorption curveof the constituent which has produced the color. In many cases, he coloris determined by substanceswhich isolated are colorlessor have a differ-ent color from that which they bestow on the minelral n which they are

found.K. Chudoba (20) has shown that greenzircons have a specificweightIess han ZrSiO+and reveal some difierences n the structure; they con-tain amorphous SiOz and ZrOz. On heating to 14500, he specific weightincreases, nd the color changes rom blue to yellow.Perhapsthe presenceof amorphous material might be due to radioac-tive substances.Machawsky (21) has drawn attention to other varia-tions of color due to the heating; morion becomes ellow at 5000;ame-thyst becomescolorlessat 300-4000, hen opalescentand then yellow at575-7500;some amethysts from Paraguay become yellow at 390-4500,but return to the amethyst color upon exposure o radium. Analogously,as Chudoba has pointed out, green beryl becomesblue if it is heated at400"; yellow or brown topaz becomespink upon heating, and with ultra-violet rays changes o yellow again.These changes of color are in part due to chemical reactions, andspecifically to a changeresulting from the oxidation of the chromophore.The changeof color of green ourmalines and of beryl when heatedwouldbe due to the oxidation of iron, that of topaz to the oxidation of va-nadium. However, in many cases, t is not a chemical but a physicalphenomenon; it is a question of the deforming influence of ions on theorbits of other ions which are easily deformed, as for example in Pbfz,which is yellow on account of the deforming influence of Pb", while NaIis colorlessbecauseNa' is only slightly deformed.


8/10

128 GUIDO CAROBBI AND KENZO PIERUCCINIIn fact, it is not possible to explain as a chemical phenomenon thepink color of beryl as due to the presenceof caesium [Klemm and WiId(22)l or the green color of the microcline as due to the presence ofrubidium and caesium(Goldschmid| (23). Likewise, according to its dis-persion, the samepigment can give difierent colors to fluorite (Chudoba,

Kleber, and Siebel) (24);so, at times, the color does not depend on thechemical nature of the chromophore.In the particular case of tourmaline, Scharizer (25) in 1889 stated forthe crystals from Schuttenhofen that the deep green color was due toiron and that the crystals became pink when manganese s present;that is , when the ratio MnO/FeO increases.When the content of ironequals manganese, he color deepenswith the increaseof the content oftitanium. The deep greenor black tourmalines are associated in Schut-tenhofen) with rutile and ilmenite, while the blue tourmalines often con-tain tin and are associatedwith cassiterite.

Scharizer (25) stated that lithium would be absent in the deeplycolored tourmalines. Rammelsberg (26) and Sommerland (27) drew at-tention to the fact that red and pink tourmalines from Scheitaukscon-tain manganesebut no iron. Riggs (28) found that the red and pinklithium tourmalines contain manganese. Kuntz (29) believes that thered and pink colors depend on trivalent manganeseand that scandiumand cerium also contribute to the coloring.MacCarthy (30) attributes yellow and green colors to iron in variousstages of oxidation. It seems hat we should pay but slight attention tothe ideas of Vernadsky (31) on the presence n the tourmaline of akaolinic nucleusthat, added to other groups, would give colors, as is thecase n sodalite, haiiynite, and cancrinite. The presenceof this nucleusdoesnot appear to be in accord with the composition of tourmaline.In the opinion of Wild (32), the chromophores n tourmaline are Fe,Mg, Mn, and Co (magnesium would not cause color, but would havesome nfluence on the others). Secondaryagenciesof color would be Gaand Sn; casual chromophores nclude Ni, Pb, Bi, and Sn. Wild assignsthe responsibility of every color to a combination of elements. In theopinion of Holden (33), the color of rubellite is due to trivalent man-ganese while the divalent and the tetravalent manganesewould not af -fect the color), and colorization by heating would be a process nvolvingreduction.Recently, T. W. 'Warner (34) made a spectrographic analysis of thetourmalines from San Diego,* and he has found the following elementspresent n everysample:Li, Na, Mg, Ca, AI, B, Mn, Si, V;and in someof

* One sample is indicated as coming from Brazil'


9/10

SPECTROGRAPHICNALYS$ OF TOARMALINESFROM ELBA 129them K, Cu, Fe, Pb, Sn, Ti, and Be. In the opinion of Warner, Li, Na,Mg, Ca, Al, and B have no influence on colorlcopper is more frequent inthe greentourmalines than in those which are pink; instead, tin is moremarked in the pink.

Concluding our researches rom the point of view of color, we mayaffirm that:The greencolor, as previously noted, is due to the presenceof divalentiron.The pink color is due to the presenceof manganesewith lithium andcaesium.The blue color is causedby the presenceof copper, and particularly ifstrengthened by the strongly deforming action of Cu/ in the same wayas in the copper ammonium complex, where the deep blue color is due,according to Fajans, to the deforming action of Cu", on the electronicorbits of the molecule of ammonium.

Rnlnnpucns(1) BnoNcxrmr, rait|deMin4ral,ogi,e,l,06,Paris 1807).(2) Tozznrn, O. T., Minerali particolari dell'isoladell'Elba ritrovati e raccolti dal Sig.GiovanniAmmannati, Florence 1825).(3) D'Acnar-or, A,.,M ineralogi.adell,aT oscana,Pisa,2, 19 (187 ).(4) D'Acro.lnor, G., Le tormaline del granito elbano:Atti ilella soc. Tosconali. ScienzeNatural,i,Pisa,Memori,e,121893), nd f5 (1896).(5) Gnrrr-, E., Nuove formecristallinedella tormalinaelbana:Atti del,l,aoc.ToscanaliSci.eweNaturol,i,Pisa, Memorie,34, 243 (1922).(6) Bunncrn, X{. J., eNoP,trlnrsr,W., The unit celland space roup oI tourmaline:Am.M'ineral., 2, ll39 (1937.(7) Kurasznwsrr, Cnanrorre, Uber die Kristallstruktur des Turmalins: Abhandt,.math.phys.Klasse. tichs. kad..Wissen.,38 1921,.(8) Macrarscmr, F., Die Formeleinheit erTurmalins:Zeits.Krist.,70,224 (1929).(9) Srnocr,L.W.,Ztr GeochemiedesLithiums:aehr.Ges.Wiss.Gdttingen,Math.-Physi.k.Kl,asse, achgruppeV, l, No. 15 (1935).(10) Goroscrnmr, V. M., BrnunN, H., HlulrlleNx, H., uxo Prtnns, Ct-., Zttr Geo-chemie der Alkalimetalle I: Nachr. Ges. Wi,ss.Gdttingen,Math.-Physik. Klasse.FachgruppeII , No. 34,W, No. 35 (1933).Gotoscrlfiot, V. M., Beunn, I{., Wrrtr, H.,Ztt Geochemie er Alkalimetalle II.,Ib,i.d.., ochgruppe\1, l, No. 4 (1934).(11) voNENcnr,rra.nor,W.,DieGeochemiedesBarium:Chemi'ei lerEri le, lO,87(1936).(12) Gor.osnuror, V. M., uxo Pnrnns, Ct.,Ztr Geochemie es Berylliums Nachr. Ges.Wiss.Gi)ttingen, achgruppeII, No. 23; fV, No. 25 (1932).

(13) Gor.oscrnnor,V. M., unn Prmns, Ct ., Zur Geochemie es Scandiums:Nachr. Ges.Wi'ss.Gdttingen, achgruppeII, No. 14; FachgruppeV, No. 16 (1931).(14) C.r.r ionnr,G.,Rend.R.Accad..Naz.Lincei,Series5,33,pp.246and322(1924).(15) Mi'n. Petrog.Mitt'., 18, 304 (1899).(16) Zeit. norg. Chemie,12,375 (1896) M in. P etro .M i t., L9, 144(1900).(17) Fortsch.Mi'n.,129 (l9ll).(18) / ' / .Jahrb.Mi.n.,B.B.57,1 (1928).


10/10

130 GUIDO CAROBBI AND RENZA PIERUCCINI(19) If. Jahrb.Mi.m.,8.8.69,183 1935).(20) Deutschen olilschnried,ee'ilurcg, o.,rA (1937).(21) Deutschen olilschmi,eileeitung,No. 23 (1939).(22) Centr.Mi.nerol,., 95 (1925).(23) Nachr.Ges.Wi.ss. dttdn:gen,I,8 (1934).(24) Chemie .erErde, 13, 472 (1941).(2,5)Zeih. Krist.,L5,337 (1889).(26) Mi.neralChern., , 540 (1872).(27) Zeits.Kryst., LL,440(1886).(28) A*t,. our. Sai. 3),35 (1888).(29) Chemie er Erile, 4, 226(1940).(3O)Am, Mineral,,ll,3zI (1926).(31) Zeits, Kryst., 53, 387(1932).(32) Centr.Min.,327 (L93r).(33) Arn.Mineral.,g,l0I (1924).(34) Am. Mineral'.,20,531(1935).

Gems Tourmaline

Documents

Transcript of Gems Tourmaline