Cicncias da Terra (UNL) Lisboa N.O13 pp. 9-2212 figs. 1999

Alteration and alterability of the anorthosite from Angola

Joaquim Simao & Zenaide C. G. Silva

Centro de Estudos Geol6gicos, Depto. Ciencias da Terra, Fac. Ciencias, e Tecno logia, Univ. Nova de Lisboa. 2825-114 Caparica, Portugal

ABSTRACT

Key wo rds: Anorthosite; dimension stone; alteration; polluted environment; acid rain.

Siliceous rocks are widely used as dimension stone but the last decades have registered an increase rate of their alterationwhen exposed to polluted environments . Anorthosites were treated by acidified solutions ofHCl, HN03 and H2S04 simulating acidrain and the response was recorded through different experimen ts such as on the surface of the polished rock and on the surfaceof uncovered thin sections. The main components, plagiocl ase and olivine, both responded in similar ways to each acid solution,although following different trends; while plagioc lase develops a thin layer which acts as protection to the mineral, olivine at firstundergoes alteratio n due to leaching of magnesium and iron and in a following stage, is mechanically removed from the rock. Theaction of warm water on the rock was tested through the use of the Soxhlet extractor which caused changes on the rock colourand leaching of several cations from the its components.

RESUMO

Palavras chave: Anorto sito; rocha ornament al; alterac ao; ambientes poluidos; chuva acida.

As rochas siliciosas sao amplamente usadas como rocha ornamental tendo, nas ultimas decadas, registado um aumento dograu de alteracao quando expostas a ambientes poluidos. ° tratamento de anortositos com solucoes de HCl, HN03 e H2S04

simulando chuvas acidas foi desencadeado de modo a permitir a observacao dos estados pregressivos de alteracao, tanto nasuperficie da rocha polida como na superficie de laminas delgadas nao cobertas. Os componentes principais da rocha, plagioclasee olivina , responderam de modo semelhante aaccao de cada solucao acida apresentando, no entanto, aspectos diferentes; enquantoa plagioc lase desenvo lve uma pelicula na superficie , que actua como proteccao do mineral, na olivina observa-se a lixiviacao deMg e Fe seguida de remocao mecanica do grao. Testou-se tambern a accao de agua quente na rocha, usando 0 extractor de Soxhlet,observando-se alteracao na cor da rocha e lixiviacao de varies catioes dos seus componentes.

2

1. INTRODUCTION

Acid rains produced in polluted environmentshave accel erated the process of rock decay byweathering and have become a serious threat to oldmonuments, known through centuries, left bymankind as relict of their culture and habits indiffe rent places and at different times in the world.Urban and industrial areas are typical examples ofthese environments where the action of acid rainsis increasingly felt , leaving behind unprecedentrecords of their effects.

The study of rock alteration requires aprevious understanding ofboth the composition and

the expected behaviour of the rock, whenever it isexposed to specific conditions. Experiments werecarried out on anorthosite which is an ideal type oflithology to study alteration, due to its simplemineralogy, as already recognised by other authors(Fritz & Mohr, 1984). The samples are from theKunene Complex, Angola, known as the "blackgranite", which due to its homogeneous blackcolour and texture has been widely used asdimension stone in Portugal and elsewhere.

After the rock had been characterized throughphysical tests, experiments were carried outenvolving the use of the Soxhlet extractor, in orderto detect the alterations due to leaching produced by

9

Cit2/1cias da Terra (UNL), 13

rain water in tropic al condi tions. Simultaneously,polis hed fragments of the anorthosite were treatedwith acidified solutions of HCI, RN03 and H2S04,

simulating the composition of acid rain water inpolluted environments. The purpose of this researchis to follow the extent and the type of alteration inthe rock, assuming that the action of the foreignagent is similar to the conditions which occurnaturally in polluted atmospheres.

2. PETROGRAPHIC, MINERALOGIC ANDCHEMICAL CHARACTERIZATION

The studied samples were collected in a quarryat Chicuatite, a local ity in the Huila District(Ango la), where a large anorthosite body whichoutcrops in southern Angola and North ern Namibiahas been cut and explored as dimension stone in theKunene Complex. The dominant lithologies areanorthosites and troctolites and both rock typeshave been commercially used for that purpose.These rocks are very dark , coarse grained andmacroscopically are very similar. The wholeComplex has been studied by Silva (1987, 1990and 1992).

The anorthosite mineralogy is simpl e, havingpla gioclase and oli vine as main components,

pyroxene as common accessory and iron ore andbiotite as minor accesso ries. Under the microscopethe rocks have a cumulate texture, as definedby Irvine (1982), where plag ioclase and olivineare cumulus and the pyroxene is usually theintercumulus. This homogeneity allows a closeobservation of the progressive alteration in eachmineral , plagioclase and olivine, along the experi­ments. Figure 1 shows the polished surface of ananorthosite sample.

Petrographic analysis indicates the rock is atroctolite, according to Streckeisen classification(1976 ). Calcic plagioclase is the main component,followed by olivine and other ferromagnesianminerals , as shown in Table I. All minerals lookfresh in appearance.

Table I. Mineralogic al compos ition.

Min erals %

Plagioclase 80.0

Olivine 15.0

Pyro xene 4.5

Biotite 0.2

Opaque (oxides) 0.3

Figure 1. Photograph of polished anort hosite in hand speci men.

10

The rock was analyzed in the laboratories ofthe Geology Department of the University ofLisbon, Portugal. Wet chemistry was used in thedetermination of silica and atomic absorption

(Varian Spectra, A-30), for the other oxides.Chemical and normative compositions are pre­sented in Table II.

Table II. Chemical and normative composition.

OxidesI

(%) Normative MineralsI

(%)

Si02 49,55 or 1,83

AI20 3 21,70 ab 25,30

Fe203 3,98 an 44,88

FeO 4,51 di 0,10

MgO 7,10 hy 18,06

CaO 9,07 01 3,27

Na20 2,99 mt 5,77

K20 0,31 Plagioclase = A I164

Total 99,21

3. PHYSICAL CHARACTERIZATION

Physical tests (Table III) were performed atthe Geothecnics Laboratory of F.C.T. (UNL),according to standard procedures specified by

RILEM (1980), as such: porosity accessible towater, real and apparent volumic mass, coefficientof water saturation, water absorption at lowpressure, expansibility by water absorption and saltcrystallization test.

Table III. Physical properties of the anorthosite.

Physical properties Average

Porosity accessible to water 0,1 %

Real volumic mass 2844 Kg/m:

Apparent volumic mass 2841 Kg/rn:

Coefficient of water saturation aprox. °%

Water absorption at low pressure ( I h) °em-

Expansibility by wate r absorption (72 h) aprox. °mm

Salt crystallization test (30 cycles) 0,18 %

These properties allow a classification of thisanorthosite as "of good quality" , suitable for thepurpose it has been used in construction.

4. EXPERIMENTAL ALTERATIONOF THE ANORTHOSITE

Two types of experiments were performed onthe anorthosite. Rock fragments were tested withthe Soxhlet extractor and uncovered thin sectionswere wiped with acidified solutions of HCI, HN03

and H2S04 , The progressive alteration experiencedby the rock was followed through microscopicobservations and chemical analysis on the rocksurface, by using the petrographic microscope andthe SEM. Variations on the chemical compositionof all liquids after leaching the rock samples werealso determined through chemical analysis.

4.1. Experiments with the Soxhlet extractor

The Soxhlet extractor is used to simulatethe effect of rain water on rocks through leaching

11

Ciencias da Terra (UNL), 13

..

//

/II'

pH 10,09,08,07,06,05,0

o 1000 2000 3000 4000 5000Hours

6000

Figure 2. pH variation at the end of each 1000 hours cycle.

o~---,;i--;i--:i:--:i:--;i------,i Hours

ooon

oo'"

oo'"

oo

'"

oo

oo

~-+--+--+--+--+--... HoursosMgO(H P ) x 100

MgO (Rock)

0 ,00100

0 ,00075

0 ,00050

0,00025

3

2

CaO (HP) x 100CaO (Rock)

---,AI.,...=P,:--:::-(H....:.P,.:..) x 100AIP j Rock) 4

0,04

_ S_iO....::i:-HP::....:-) x 100

SiOiRocha) 0 ,12

~::~~0 ,0006

0 ,0003 ~ _

o Hoursa 0 a 00 0o 0 0 0 a 0- N ~ 7 ~ ~

Nap+Kp (H P) x 100Nap+Kp (Rock)

2,0

1,5

1,0

0 ,5

oW"-- +-- +-- +-- +--+_------,I

Fig. 3. Cumulate (-0-) and absolute (-+-) varia­I tions of SiOz,Alz0 3, MgO, CaO and Na.O + KzO

during the Soxhlet experiment.

0 ,08

Table IV. Solid residues from the Soxhlet experimentafter each 1000 hours cycle.

Hours Products

1000 silica gel; feldspar ; zeolite ?2000 silica gel; kaolinite3000 crystalline kaolini te4000 crystalline kaolinite5000 kaolinite6000 poorly crystallized kaolini te

during a specified period of time. Similar experi­ments have been performed in other rock types byAires-Barros (1971, 1972 and 1991) , in Aires­Barros & Mouraz-Miranda (1979) and Mouraz­Miranda (1984 and 1986), where the methodologyis described.

The experiment was performed during sixcycles of 1000 hours each , in a total of 6000 hours.At the beginning of each cycle the distilled waterhad a pH of 5.6, reaching 9.0 at the end ; thepercolated water was then collected and analyzed.Figure 2 shows the pH variation during theexperiment and shows that although there is asubstantial difference between the beginning andthe end of each cycle, the range of variation is notsignificant among cycles. Estimation of the pH forthe solutions was performed in a SOTEL WTW pH350 , at the Chemical Engineering Dep artment of1ST (Technical University of Lisbon).

Chemical analyses for major elements wereperformed in waters at the end of each cycle, bycollorimetry and spectrophotometry, in a VarianSPECTRA A-30 at the GeoFCUL laboratories. Therecorded data are presented in Fig. 3.

Solid res idue left after filtering each solutionwas analyzed by X-ray diffraction (Phillips PW1050170), at the Geosciences Department of theUniversity of Aveiro . Table IV reports the productswhich were identified at the end of each cycle(silica gel , feldspars, zeolites and kaolinite).

12

2 052

15 0.4

03g I 0

"'02

05n.1

0 0

lnicio /0 3° 4° 5° 6° 7u Infcio 1° 2° 3° 5° 6°

Figure 4. Weight loss cumulate variation of the anorthosite samples after treatment with acidified solutions ofHCl (- -), H2S04 (- -) and HN03 (- -) . (Graphic 1: samples 4 x 4 x 0.7 em; Graphic 1: samp les2 x 1.5 xO.5 em).

Most alterations in the anorthosite occurduring the first two cycles, indicating a highercation mobility during that stage . The increase inthe oxides contents in water, shown in Fig. 3illustrates this fact. After the 2000 hours stage , thevariations in the concentrations are similar andall oxides register a decrease in their contents.The concentration in alumina though , after the twofirst cycles, of higher mobility, remains almostinvariabl e, as registered by the presence ofkaoliniteas the only mineral present in the rock residue.Different behaviour is shown by iron, whichexhibits a decrease in its content in the rocksamples after the 6000 hours , contrasted to itsabsence in the solution after each 1000 hours cycle.

Change in colour of the rock is a strikingfeature observed while experiments were per­formed. From an initially dark, homogeneous blackcolour, the rock changed progressively to abrownish , spotted rock, suggesting to be residualiron oxide, due to the alteration of olivine.

4.2. Treatment with chemical solutions

The effect of chemical weathering activatedby acid rain in the anorthosite was studied bypromoting the treatment of rock samples withacidified solutions, simulating a natural situation.Polluted atmospheres carry sulphur, nitrogen andchlorine which are the common gases responsiblefor the production of acid rain, as they combine orare dissolved in water, lowering he pH of thenatural rain water (around 5.6 in the presence ofCO2 and atmospheric pressure) . Taking that intoaccount, the experiments were performed with HCI,HN03 and H2S04 , The solutions were prepared bydiluting (0.25% (v/v)) the concentrated acids assuch: HCL (37%) and HN03 (65%) Prolab; H2S04

(98%) Merck; the pH in the final solutions were1.16, 1.41 and lAO for HCI, .HN03 and H2S04

respectively.Three slabs of4 x 4 x O.7 ern (ref. 1, 2 & 3) and

18 others of2 x 1.5 x O.5 em (1.1 to 3.6) were cutfrom the polished rock. Three sets of small sixsamples each were dumped in 80 ml of eachacidified solution and every 15 days one of themwas removed . At this point, the solution wasanalyzed as it was also the surface of the rock . Theweight loss for each sample was then recorded; thetotal , cumulative loss is illustrated in Fig. 4. Majorloss in weight occurred during the first two cyclesand showed to be more effective in samples treatedwith H2S04, as already recognized by Martin et al.(1992) . Weight loss in smaller samples are similarduring the four initial cycles when treated by HCIand H2S04; from there on, this last acid is moreeffective; HN03 is the least effective of all threecomponents.

Major element analys is by atomic absorptionwere performed in a PERKIN-ELMER spectro­meter (model 2380) at the Earth ScienceDepartment ofF.C.T. (UNL) . Concentrations ofCa,Mg, Na, K, AI, Si and Fe were determined byspectrophotometer and the corresponding inter­ferences were eliminated according to EAA(Perkin-Elmer, 1982). The evolution of the con­centration of major elements in the solutions isillustrated in Fig. 5.

The action of different acids on the sampleswas slightly different, but in a general way, themobility of all elements, as seen from the graphics,is higher during the first two cycles, regardless ofthe sample size.

As silica is concerned, the HN03 solution isthe least effective of the three acids when used onsmaller samples (Fig . 5.7); HCI displaces alum ina

13

Ciencias da Terra rUNL), 13

more easily than other acids, and its action does notdepend on the sample size.

Iron and magnesium seemed to be the mosteas ily removed cations, due to the alteration anddes-agregation of olivine; the action of all acids wasstronger in the larger samples. (Fig. 5.3 and 5.4) .

Calcium was removed more effectively by HCI;however, gypsum crystals were left in the powdereddry residue from the H2S04 solution, indicating thatthis cation did not remain in this solution. The atomicabsorption low signal given by calcium in sulphuricmedium could also be accounted for the low calciumcontent registered in that solution.

Sodium and potassium were recorded togetherdue to their low concentrations in all solutions.

The action of different acids are illustrated in Fig.5.6 and 5.12 .

Observation of hand specimens after treatmentwith acidified solutions

Hand specimens observations indicate that theeffect of all acidified solutions was similar; therock , from init ially black, turned to a whitish, roughmaterial, exhibiting small cavities on its surface,corresponding to the location of the olivine grains;at first they were chemically altered, followed by amechanical removal. Figure 6 illustrates the twostages, prior and after the use of H2S04, (7-15 dayscycles).

~~I~' ; Tre atments

1 2 3 456 7

AI (ppm/g)

2 3 4 5 6 7

:~: 0 Treatments

1 2 34 5 6 7

52 3 43 4 5 6 7

:~~: ~~--I -:-Trcal l11en ts6 7

Fe (ppm/g)

I6

:~IO; 6 :::::=:=::::::::::::::::::::~ ~-+I--t----II Trea tments

I 2 3 4 5 77°:

6°I

5°

:~ t ~~4 ===-=

~te: IJ" 2" 3" 4"

Mg (ppm/gl

C. (ppm/g)

~k:i:=:2 3 4 5 6 7

N.+ K (ppm/g)

2 3 4 5 6 7

:~1~~Tre.tmcnts

12 3 45 6 7

Figure 5. Cumulate vanation of major elements on the solutions after treatment of anorthosite samples withHCI (- -), H2S04 ( - -) and RN03 ( - -). (Graphics I to 6: samples 4 x 4 x O.7 em; Graphics 7 to 12:samples 2 x 1.5 x 0.5 em) .

14

Figure 6 - Fresh anorthosite sample (I) and altered (2), after 7 treatments (15 days each) with H2S04 (0,25%).

Observation of thin sections after the use ofacidified solutions

As previous ly referred, three uncovered thinsections of the anorthosite were wiped wi thacidified sol ut ions in order to allow a dai lyobservation of the progressive alteration. The effec ton the surface of the rock was recorded afte r every15 days cycles and the changes were compared foreach acid . Figures 7, 8 and 9 are illustrations ofthese changes.

The response of the anorthosite to the acidtreatment was simi lar for the different solutionsalong the three cycles: progressive alteration ofolivine until its complete removal and grad ualsharpening of the fractures and cleavages of otherminerals, particularly of the pagioclase grains.

Simao & Silva (199 5), Silva & Simao (1995)and Simao (1996) have reported some particularfeatures observed in the olivine grains under themicroscope: gradual changes in the interferencecolour, reflecting the va riat ions in chemicalcomposition and variations in the appearance of thegrain, up to its total removal after the seco nd cycle.These grai ns were repl aced by a ye llowish,unidentified material, possibly limonite. Similarobservations are reported by Fritz & Mohr (1984)in olivines of close composition to the ones studiedpresentl y. It should be mentioned that after the first

cycle the olivine grains treated by HCI were gone,emphasising the severity of its action on thatmin eral (Fig. 7).

The other mineral components, namelypyroxene and plagioclase, showed to be moreresistant; their alteration consisted mainly on thesharpening of their grain contours and on thecleavage traces and on the gradual darkening of thegrains, similar to that observed in a saussuritizedpagioclase.

Observation and chemical analysis of the surf aceof the rock samples

After the treatment with the acid solutions, thesurfaces of all samples were analyzed using ascanning microscope so that the progressivealteration of all minerals could be detected.

The microscope was a l EOL 330A (0,6 A, 20kv), connected to a TRACOR (ED S) Microprobe.Semiquantitative chem ical analyses were restrictedto the two main components of the anorthosite,plagioclase and olivine, and recalculated to 100%.Due to the fact that the analyses are referred to anarea and not to a point, each one might includesom e trac es of neighbouring minerals. The evolu­tion of the analyses of samples 1.1 to 3.6 areregistered in Figures 10 and 11.

15

Ciencias da Terra (UNL), 13

fresh roch

2nd Cycle

1st Cycle

3rd Cycle

Figure 7. Evolution of the anorthosite alteration observed in thin section after treatment with a 0.25% Hel solution(01- olivine; pi - plagioclase; nic6is X; f----1= 0.1 mm).

16

fresh roch 1st Cycle

':.....

2nd Cycle 3rd Cycle

3

Figure 8. Evolution of the anorthosite alteration observed in thin section after treatment with a 0.25% H2S0 4 solution(01 - olivine; pI - plagioclase; nic6is X; f----1 =0.1mm).

17

Cit2ncias da Terra rUNL), 13

fresh roch

2nd Cycle

1st Cycle

3rd Cycle

Figure 9. Evolution of the anorthosite alteration observed in thin section after treatment with a 0.25% HNO~ solution(01- olivine; pl- plagioclase; nic6is X; f-----1 =0.1 mm).

18

I4

I I2 3

~ S() 2':v =-='-""""'==--======40

t20

o :-;,- -l-- - +-- -+--+---+I- --IITrea tmentsallt;lUit, 5 6

:~ Treatment,4 :; 6

6

I :=: ~ Trea tments

5

i5

'I ""4

,3

"~ :; r ~o~~r==~:!.'!!'!Sj:~~~~=~1 T reatments

rd, I 2 6l\llf I lla t

5I/\" ''''' 1~ot I I j Iallt~;:'t' 1 2 J 4

'c~ :~l ~o~ I ::s;r- , I <:::: I=--==; Treatm ent s

;U1(~~\~ t' I 2 :1 4 5 6

~ ....::;::::::I · - i I I Treatments

~Treatmen t':; 6

2 3 4 5 6

Figure 10. SEM major elements analyses of plagioclases performed on the anorthosit e surface after treatment withHel (- - ), H2S0 4 (- - ) and HN03 (- -) .

The response of plagioclase was similar in allsamples, reg ardless of the nature of the acidsolution. There is a relative increase in the silicacontent and decrease in the othe r components,observed on the grain surface. The apparentincrease in MgO and FeO at the end of the firstcycle is interpreted as due to the contribution of theneighbouring oli vine, Oli vine had a similarbehaviour, exhibiting a superficial increase in thesilica content and decrease in the other components,registering the lower reactivity of RN03 ascompared to the HCl and H2S04 solutions (Fig . 11).

Berner & Holdren (1979) have pointed outthat the mechanism which rules the alteration anddissolution 'of plagioclase is a superficial reaction.

In fact , Busemberg & Clemency (1976) haverecognized that initially the dissolution is promotedby the exchange of the mineral cations and thehydrogen from the solution; this hydrogen fromaci d solutions is able to percolate trough theplagioclase surface and infiltrate down to hundredsof angstroms (Casey et a!., 1987, 1989a, 1989b).

Figure 12a illustrates the changes occurred onthe surface of the plagioclase gra ins, as regularfractures developed after the thin layer of hydratedamorphous silica had dried out.

Casey et a!. (1988, 1989a, 1989b) havedescribed a simi lar feature developed in theplagioclase during experiments carried out withsolutions having different pH; after the reaction

19

Ciencias da Terra rUNL), 13

~~ Treatments2 3 4 5 6

65

==I I T reatments5 6

43

-c::::=I T3 4

5O~"' h.-O

40

30r~.--:::::::::::.". "~~tSc: ~=-::§::--§- [Tr eatmen ts

~~l: 1 250","'?I~K:=>~~~l' I 2

s::;> ?;;s;=~I ; I : Treatments2 3 4 5 6

"""~,,,, ;~32

I . ======01 I~I~ Treatments

ill:~~t' ] 2 3 4 5 6

Figure II. SEM major elements analyses of olivines performed on the anorthosite surface after treatment withnet (-), H2S04 (-) and HN03 (-).

with solutions of pH lower than 4, the mineralsurface is gradually enriched in silica, depleted inalumina, calcium and sodium, causing an increasein the porosity and widening the mineral structure;with a pH close to 1, after several hundreds ofhours of experiment, a thin layer (nm thick) ofrepolimerized silica develops on the mineralsurface, as already seen on Fig. 12a.

Some alteration develops near plagioclasecleavages and fractures, although the chemicalcomposition of the mineral remains apparentlyunchanged. In Fig. 12b these altered areascorrespond to the light zones. Gypsum crystals areadditional features observed on the plagioclasestreated with H2S04 , as illustrated in Fig. 12c.

Alteration of olivine is similar in all situations;the corrosive action over the grains surfaces is

intense, after the treatment with the solutions. Atthe end of the complete cycles the grains weretotally destroyed, except their framework and theirfractures, presently filled with more resistantmaterial (Fig. 12d). Olivines consume hydrogenfrom the acidified solutions and releasesmagnesium, iron and silica to the solutions (Nesbit& Wilson, 1992). Variations in the iron contentshown in the curves of Fig. II are possibly due toretention of the hydrated iron oxide on the mineralsurface during the alteration process. Con­centrations of AI, Ca, Na e K as recorded in theolivine analysis can be accounted for the highmobility of these cations in the solutions and to theparticipation of the neighbouring minerals in theanalyzed surface.

20

Ciencias da Terra (UNL), 13

These data give a useful clue to the changes ofminerals in polluted environments and as suchcould be used as indicators of the behaviour of

REFERENCES

the silicate rocks m previously known pollutedatmospheres.

Aires-Barros, L. (1971) -Alterac;ao e alterabilidade de rochas igneas. LNEC, Lisboa.

Aires-Barros, L. (1972) - Analise laboratorial da influencia da granu1aridade na alterabilidade das rochas. Tecnica, 417: 65-70.

Aires-Barros, L. ( 1991) -A1terac;ao e a1terabi1idadede rochas. INIC, Lisboa.

Aires-Barros, L. & Mouraz-Miranda, A. (1979) - A alteracao laboratorial de rochas pelo extractor de Soxhlet e sua repercussao naoxidacao Fe2+ ~Fe3+ e na variacao da densidade aparente. Com. Servo Geol. Portugal, LXIV: 243-256.

Berner, R.A. & Holdren, G. R. (1979) - Mechanisms offeldspar weathering - II. Observations offeldspars from soils. Geochim.

Cosmochim. Acta, 43: 1173-1186.

Busemberg, E. & Clemency, C. V. (1976) - The dissolution kinetics offeldspars at 25°C and 1 atm CO2 partial pressure. Geochim.

Cosmochim. Acta, 40: 41-49.

Casey, W. H., Westrich, H. R. & Arnold, G.w. (1988) - Surface chemistry oflabradorite feldspar reacted with aqueous solutionsat pH = 2,3 and 12. Geochim. Cosmochim. Acta, 52: 2795-2807.

Casey, W. H., Westrich, H. R., Arnold, G. W. & Banfield, J. E (1989a) - The surface chemistry of dissolving labradorite feldspar.Geochim. Cosmochim. Acta, 53: 821-832. .

Casey, W. H., Westrich, H. R., Massis, T., Banfield, 1 E & Arnold, G. W. (1989b) - The surface of labradorite feldspar after acid

hydrolysis. Chem. Geol., 78: 205-218.

Fritz, S. 1 & Mohr, D. W. (1984) - Chemical alteration in the micro environment within a spheroidally-weathered anorthositeboulder. Geochimica et Cosmochimica Acta, 48: 2527-2535.

Irvine, T. N. (1982) - Terminology for layered intrusions. J. Petrology, 23(2): 127 162.

Martin, L., Bello, M.A. & Martin, A. (1992) - Accelerated alteration tests on stones used in the Cathedral of Granada (Spain).7th International Congress on Deterioration and Conservation a/Stone, LNEG, Lisboa, pp. 845-850.

Mouraz-Miranda, A. (1984) - Alteracao experimental de rochas - a contribuicao do Instituto Superior Tecnico. EstudoComplementar de Dissertacao de Doutoramento, Instituto Superior Tecnico, Lisboa.

Mouraz-Miranda, A. (1986) - Alteracao experimental de rochas - a contribuicao do Instituto Superior Tecnico, Geotecnia, 48:

3-37.

Nesbitt, H. W. & Wilson, R. E. (1992) - Recent Chemical Weathering of Basalts. America Journal a/Science, 292: 740-777.

Perkin-Elmer, (1982) - Analytical Methods for Atomic Absorption Spectrophotometry, Norwalk.

RILEM - Reunion Internationale des Laboratoires d'Essais et de Recherche sur les Materiaux les Constructions. Commission 25- PEM Protection et Erosion des Monuments (1980) Essais recommendes pour mesurer l'alteration des pierres et evaluerl'efficacite de methodes de traitement. Materiaux et Construtions, Paris, 13(75).

Silva, Z. C. G. (1987) - Estudo Petroquimico do Complexo Gabro-Anortositico de Angola. Ph. D. Thesis, Univ. de Lisboa,Portugal.

Silva, Z. C. G. (1990) - Geochemistry of the Gabbro-anorthosite Complex of southwest Angola. 1. Afr: Earth Sci., 10: 683-962.

Silva, Z. C. G. (1992) - Mineralogy and cryptic layering of the Kunene anorthosite complex ofSW Angola and Namibia. Mineral.

Mag., 56: 319-327.

Silva, Z. C. G. (1995) - Petrologia, Geoquimica e Potencial Econ6mico de um Corpo Igneo Estratiforme. Relat6rio final, JNICT,

Lisboa.

Silva, Z. C. & Simao, 1 A. R. S., (1995) - Alteracao Progressiva em Rochas Silicatadas. Ensaios em Anortositos. Mem6riasn.? 4, IV Congresso Nacional de Geologia., Fac. Ciencias, Univ. Porto, Museu e Laborat6rio Mineral6gico e Geol6gico,Porto.

Simao, J. A. R. S. & Silva, Z. C. G. (1995) - Composicao Mineral e Alteracao de Rochas Silicatadas.o "Granito Negro de Angola" - Um caso em Estudo. 1.0 Congresso Internacional da Pedra. ELL., Lisboa.

Simao, lARS. (1996) - Alteracao e alterabilidade do anortosito de Angola utilizado como rocha ornamental. Internal report,

Faculdade de Ciencias e Tecnologia, Universidade Nova de Lisboa.

Streckeisen, A. (1976) - To each plutonic rock its proper name. Earth Sci. Rev., 12: 1-33.

22