analyzing growth kinetics of magmatic crystals by backscattered

Analyzing growth kinetics of magmatic crystals by backscattered electron

microscopy of oriented crystal sections

R. Sturm1

1 Brunnleitenweg 41, A-5061 Elsbethen, Salzburg, Austria

In the past decades scientific studies on crystal growth were routinely carried out by the application of light-microscopic

methods and, later on, by the production of oriented crystal sections and their documentation using electron-microscopic

imaging techniques (BSEI, CL, etc.). Based upon numerous microscopic investigations published hitherto, the magmatic

mineral zircon (ZrSiO4) is characterized by a wide spectrum of growth trends, whereby three main crystal developments

are distinguished here: (1) Undisturbed growth of zircon exhibits a periodical development of the two F-forms {100} and

{101} without any change of the grain morphology during the entire crystallization process. (2) Transitional (inhibited)

growth of zircon is marked by a successive replacement of the F-forms {100} and {101} by the S-forms {110} and {211}

or {311}, resulting in moderately elongated grains with steep pyramids. (3) Untypical or "pathological" growth of zircon

shows a contrary trend with respect to transitional growth, i.e. the primarily occurring S-forms {110} and {211}/{311} are

continuously substituted by the F-forms {100} and {101}, resulting in long-prismatic crystals with flat pyramids.

Keywords back-scattered electron microscopy; crystal growth; growth rates; growth control; zircon

1. Introduction

Since the pioneering mineralogical and crystallographic studies of the 1960s and 1970s it has been regarded as a main

fundamental of earth sciences that growth of minerals and especially of magmatic crystals depends upon the physico-

chemical state of the growth environment. Based on extensive empirical and experimental approaches, main factors

controlling crystal growth and, as a consequence of that, crystal morphology could be detected and documented

meanwhile [1-3]. Therefore, chemistry and temperature of the growth environment as well as its content of volatile

phases have been found out as most essential in the overwhelming number of cases of crystallization [4, 5].

Concerning the crystallization of mineral phases from a magmatic source, accessory zircon has attracted most

geoscientific attention in the past decades [6-8] due to several reasons. First, zircon is characterized by a somewhat

ubiquitous occurrence, i.e., it can be separated from magmatic, metamorphic, and sedimentary rocks. In very

exceptional cases, the accessory mineral can be also detected as a component of extraterrestric rocks. Second, zircon

exhibits a strong resistance to any processes in the earth’s crust, thereby often being involved into several cycles of

rock-forming processes [6]. A third reason underlining the scientific exclusiveness of zircon is its development of well

describable and easily distinguishable crystal shapes which, in most cases, are composed of two prisms and two

pyramids (biprismatic bipyramidal morphology).

A systematic empirical description of the various zircon shapes occurring in igneous rocks was carried out in the

1970s and 1980s, leading to the introduction of the so-called typology diagram [7, 8]. This represents a graphical

scheme containing all zircon morphologies described hitherto as a combination of the prism forms {100}, {110} and the

pyramidal forms {101}, {211}/{311}. As decoded by an extensive statistical evaluation of different zircon populations,

the predominance of certain forms mainly depends upon the chemical and thermal characteristics of the magmatic

source, i.e., Al-rich magmas with lower temperature usually develop other crystal shapes than alkaline magmas with

higher temperature. Hence, zircon was evaluated as a petrogenetic indicator bearing basic information of its source rock

and even as a geothermometer having recorded the temperature during crystallization [8].

Whilst from the 1960s to 1980s zircon morphology was mainly studied using various techniques of light-microscopy

[8-10], from the 1990s on electron-microscopic methods made their increased appearances [11, 12]. These innovative

and partly even revolutionary methodical approaches provided a direct insight into the crystal development, leading to a

new understanding of the essential relationship between zircon growth and chemistry of the magmatic source [13].

Nowadays, electron microscopic analyses including BSEI and CL have to be regarded as basic procedures to obtain

advanced information concerning the important role of accessory zircon as petrogenetic indicator.These methods, how-

ever, require an appropriate preparation of the zircon crystals, including the production of polished sections which are

oriented either parallel or perpendicular to the crystallographic c-axis and therefore enable the elucidation of chemical

factors controlling the morphological development of pyramidal and prismatic faces [14, 15].

Microscopy: Science, Technology, Applications and Education A. Méndez-Vilas and J. Díaz (Eds.)

©FORMATEX 2010 1681

______________________________________________

2. Preparation and analysis methods

2.1 Zircon crystal preparation for electron-microscopic studies

Microscopic analysis of zircon growth was preceded by an appropriate preparation of single zircon crystals. Depending

upon whether prism growth or pyramidal growth was studied, cross or longitudinal sections of selected crystals were

produced according to a well-defined preparation technique [12-14]. The first stage of this procedure included the

mounting of zircon grains on a glass slide using highly viscous epoxy resin (e.g. Köropax 439). For each crystal the

crystallographic orientation was determined according to the method introduced by Frasl [10], where the apparent

termination angle of the {101}-pyramid is measured under the light-microscope. After hardening of the resin and

covering the whole preparation with an additional resin layer (thickness: 2 mm), the grinding and polishing process was

started using SiC-powders with different grain sizes as well as diamond pastes (grain sizes: 3 µm and 1 µm). In the case

of longitudinal crystal sections the thickness of the resin layer including the grains was carefully reduced by the

grinding and polishing procedure. In the case of cross sections the grains were exactly cut in the middle, and the

remaining half was subsequently mounted on another glass slide and subjected to the final polishing process. The

surface of the sections had to be polished perfectly to guarantee ideal analyses with the microprobe (Fig. 1).

Fig. 1 Preparation of zircon crystals for growth analysis with the electron microprobe. Selected grains are cut either perpendicular

or parallel to their crystallographic main axes, depending upon whether pyramidal or prismatic growth is investigated.

2.2 BSEI of oriented crystal sections and measurement of crystal growth increments

Backscattered electron imaging, the technical setup of which is schematically illustrated in Fig. 2a, is based upon the

principle that the number of electrons reflected from a sample correlates with the atomic mass number of elements in

the sample. Hence, bright growth zones of zircon on the monitor indicate higher concentrations of elements with

enhanced atomic mass and vice versa (Fig. 2b). For an optimal visualization of zircon crystals sectioned perpendicular

or parallel to their crystallographic main axes, an accelerating voltage of 15 to 20 kV, a beam current of 30 to 40 nA and

a beam diameter of about 1 µm were selected. Studies were conducted on a Jeol JXA-8600 electron microprobe.


1682 ©FORMATEX 2010

______________________________________________

a) b)

Fig. 2 a) Technical setup necessary for backscattered electron imaging of zircon crystal sections. b) Image of a zircon crystal with

its typical pattern of dark and bright growth zones (bar: 50 µm).

Pyramidal and prismatic growth of zircon crystals from high-K calcalkaline I-type rocks, peraluminous S-type rocks,

as well as highly fractionated A-type and I-type rocks was investigated by applying the quantitative growth analysis

methods formerly introduced by Vavra [16]. In general, the morphology of a crystal reflects the relative growth rates of

different faces. High growth rates (i.e., large growth increments per time) of a face will cause the gradual disappearance

of that face during growth, whilst low growth rates (i.e., small growth increments per time) can produce a significant

predominance of a crystal face over the other faces.

Concerning accessory zircon, growth increments ri of the major pyramidal forms {101} and {211} as well as the

essential prismatic forms {100} and {110} were measured on appropriate BSEI photographs of longitudinal and cross

sections (Fig. 3). Total growth of an arbitrary crystal form {hkl}, rtot{hkl}

, was simply defined by the equation

rtot{hkl}

= r1{hkl}

+ r2{hkl}

+...+ rn{hkl}

. (1)

In Eq. (1) r1...rn denote the widths of the 1st to n

th growth increment. The number of such growth bands is highly

variable in zircon crystals and depends upon the frequency, with which chemical and physical changes take place in the

magmatic source. In order to elucidate eventual changes of the zircon morphology during crystallization, growth rates

of the two pyramidal forms {101} and {211} and the two prismatic forms {100} and {110} were referred to each other

by forming the mathematical terms ri{211}

-ri{101}

and ri{110}

-ri{100}

. These differences were plotted against the distance of

the respective growth band from the crystal centre. Growth trends were obtained from the slope of a regression line

fitted through the data points.

a) b)

Fig. 3 a) Measurement of crystal growth increments ri on zircon cross sections to explore prismatic development. b) Measurement

of crystal growth increments ri on longitudinal sections to explore pyramidal growth.

2.3 EMPA to elucidate the influence of zircon chemistry on crystal development

In order to determine possible influences of crystal chemistry on the morphological development of accessory zircon,

chemical profiles using the wave-length-dispersive system (WDS) were measured. Orientation of these profiles depends

upon the section type, the intensity of growth zoning, and the possible content of an inherited core, i.e., a pre-magmatic

crystal fragment serving as nucleus for crystallization. X-rays of the chemical elements of interest (Zr, Si, U, Th, Hf, Y,



______________________________________________

Yb, P, Ca) are filtered using LIF (lithium fluoride), TAP (thallium acid phthalate) and PET (pentaerythritol) crystals.

Element quantification was enabled by the use of oxide standards of those elements named above. Respective

specifications regarding x-ray lines, standard compositions, and measurement accuracy are listed in Table 1.For the

measurement procedure on the Jeol JXA-8600 electron microprobe an accelerating voltage of 15 kV as well as a beam

current of 30 nA were selected. Correction of the raw analyses was carried out by the application of a system-internal

ZAF-4 procedure provided by Jeol.

Table 1 Specifications used for the analysis of zircon chemistry.

Oxide Crystal x-ray line Standard Accuracy (3σ)

SiO2 LIF Kα1,2 1st o. SiO2 0.135

ZrO2 PET Lα1 1st o. ZrSiO4 1.512

HfO2 TAP Mα 1st o. ZrSiO4 0.141

UO2 PET Mα1 1st o. UO2 0.164

ThO2 PET Mα1 1st o. ThO2 0.172

Y2O3 PET Lα1,2 1st o. Y2Al5O12 0.155

P2O5 PET Kα1 1st o. Ca5(OH,F)(PO4)3 0.055

Yb2O3 PET Mα 1st o. YbF3 0.242

CaO LIF Kα1 1st o. CaSiO3 0.058

3. Results of zircon growth analysis using BSEI

3.1 BSEI of oriented crystal section

Only in rare cases growth of zircon is defined by ideal geometric growth zones, allowing an exact determination of the

proportions of adjacent crystal faces at any stage of crystallization. Usually, zircon being separated from igneous rocks

is characterized by very irregular growth bands, whose development is additionally affected by phenomena such as

crystal corrosion and recrystallization. If an inherited core is included, crystal growth can be often subdivided into two

clearly distinguishable phases, an older central phase and a younger outer phase (see also crystal of Fig. 2b). The non-

ideal development of crystal morphology, however, makes a successful analysis of growth bands more difficult and

therefore requires a high extent of crystallographic experience.

Concerning pyramidal growth of accessory zircon, longitudinal crystal sections like those exhibited in Fig. 4 were

subjected to an extensive growth-band measurement. All zircon grains depicted are characterized by increased

geometric complexity due to 1) the partly very fine patterns of growth increments and 2) the appearance of several

morphometric changes within a single crystal. Comprehensive study of the BSEI-photographs generally allows a

distinction of three main pyramidal developments:

In the first case growth of the pyramidal F-form {101} and, if present, the pyramidal S-form {211} is marked by a,

let’s say, constant or undisturbed progress (crystals 1 and 3 in Fig. 4). This means that the pyramid(s) formed at the

initial stage of crystallization and being represented on the sections by the innermost growth zones retain their relative

size(s) throughout the entire growth process. Hence, the final and outermost crystal morphology is equal to the initial

and innermost morphology. Such a pyramidal growth type is chiefly observed in high-K calcalkaline I-type rocks.

In the second case growth of the pyramid {101} is dramatically accelerated during the crystallization process,

causing a continuous and somewhat passive enlargement of the pyramid {211} (crystals 2 and 6 in Fig. 4). Whilst the

innermost pyramidal morphology is characterized by the predominance of the flat pyramid, the outermost crystal shape

is chiefly defined by the steep pyramid. This inhibited growth type is commonly assigned to peraluminous S-type rocks.

In the third case described here the initially occurring pyramid {211} is subject to an accelerated growth, resulting in

a successive enlargement of the pyramid {101} (crystals 4 and 5 in Fig. 4). This so-called “pathological” growth exactly

represents the contrary case of inhibited growth and thus shows a final predominance of the flat pyramid. The

“pathological” growth type most frequently occurs in highly fractionated A-type and I-type magmas.



______________________________________________

Fig. 4 BSEI photographs of longitudinal zircon crystal sections elucidating the development of the pyramidal forms {101} (closed

square) and {211} (open square). 1, 3: zircon crystals from high-K calcalkaline I-type rocks, 2, 6: zircon crystals from peraluminous

S-type rocks, 4, 5: zircon crystals from highly fractionated A-type and I-type rocks. Bars: 30 µm.

Regarding the morphological development of zircon prisms, cross sections of single crystals like those illustrated in

Fig. 5 were studied in detail. Usually, measurement of prism growth may be evaluated as rather simple with respect to

pyramidal growth analysis due to a better and sharper appearance of single growth zones. Similar to pyramidal growth

analysis, three types of prism development may be distinguished in the study presented here:

In high-K calcalkaline I-type magmas an undisturbed growth of the prismatic F-form {100} can be recognized, i.e.,

those prism faces formed at the initial stage of crystallization retain their relative sizes throughout the entire growth

process (crystals 3 and 6 in Fig. 5). Zircon crystals characterized by an undisturbed prism development often exhibit a

significantly reduced number of growth increments, reflecting lower frequencies of chemical changes in the magmatic

source. At final stages of crystallization, processes of magma fractionation may cause an abrupt modification of the

outermost grain morphology.

In peraluminous S-type magmas there also exists an initial {100} prism showing enhanced growth rates with respect

to the {110} prism (S-form) throughout crystal growth (crystals 1 and 5 in Fig. 5). The result of this transitional or

inhibited growth is a partly complete change of the prismatic morphology with predominance or even exclusive

existence of the {110} prism in the outermost growth zones.

In highly fractionated A-type and I-type magmas a prismatic development being contrary to that described for

peraluminous S-type rocks takes place, because in this case an initial {110} prism is continuously replaced by the {100}

prism (crystals 2 and 4 in Fig. 5). Since an initial S-form is successively replaced by an F-form, this growth is again

termed “pathological”, thereby demonstrating the relationship with the respective pyramidal growth documented above.



______________________________________________

Fig. 5 BSEI photographs of zircon crystal cross sections elucidating the development of the prismatic forms {110} (closed circle)

and {100} (open circle). 3, 6: zircon crystals from high-K calcalkaline I-type rocks, 1, 5: zircon crystals from peraluminous S-type

rocks, 2, 4: zircon crystals from highly fractionated A-type and I-type rocks. Bars: 20 µm.

3.2 Results of zircon growth-band analysis

In order to obtain higher accuracy with regard to pyramidal and prismatic growth of accessory zircon, an extensive

growth-band analysis according to the methods introduced above was performed. Concerning the development of the

pyramidal forms {101} and {211} different trends of ri{211}

-ri{101}

were found for the three studied rock types (Fig. 6a):

In high-K calcalkaline I-type rocks ri{211}

-ri{101}

may be evaluated as rather constant with increasing distance of the

analyzed faces from the crystal centre, underlining the undisturbed growth already documented with the help of BSEI.

In peraluminous S-type rocks ri{211}

-ri{101}

follows a negative trend, reflecting the continuous replacement of the pyramid

{101} by the pyramid {211}. Thereby, the change from {101}>{211} to {101}<{211} takes place at early to

intermediate stages of crystal growth (intercept of trend line with the 0-line). In highly fractionated A-type and I-type

rocks ri{211}

-ri{101}

exhibits a positive trend, indicating the successive replacement of the pyramid {211} by the pyramid

{101}. The change from {101}<{211} to {101}>{211} takes place intermediate to late stages of crystal growth.

Regarding the development of the prismatic forms {100} and {110}, similar trends as obtained for pyramidal growth

could be derived from the analytical procedure (Fig. 6b). Whilst in high-K calcalkaline rocks the trend of ri{110}

-ri{100}

may be evaluated as perfectly constant, underling the undisturbed prism growth depicted in Fig. 5, in peraluminous S-

type rocks ri{110}

-ri{100}

again performs a negative trend, where the prism {100} is continuously replaced by the prism

{110}. As a main difference to pyramidal growth, the change from {100}>{110} to {100}<{110} is observed at later

stages of the crystallization process. In highly fractionated A-type and I-type rocks ri{110}

-ri{100}

again follows a positive

trend, which means that the prism {110} is successively replaced by the prism {100}. Similar to pyramidal

development of the same rock type, the switch from dominating {110} to dominating {100} is commonly recognizable

at earlier stages of crystal growth.



______________________________________________

a) b)

Fig. 6 a) Growth-band measurement and related trend analysis for growth of the pyramidal forms {101} and {211}. b) Growth-band

measurement and related trend analysis for the development of the prismatic forms {100} and {110}. Depending upon the studied

rock very specific trends of ri{211}-ri

{101} and ri{110}-ri

{100} can be recognized, remarkably underling the results obtained from BSEI

analysis.

4. Measurement of chemical profiles with WDS analysis

Measurement of chemical profiles was so far limited to cross sections of pre-selected zircon crystals from both high-K

calcalkaline I-type rocks and peraluminous S-type rocks (Fig. 7). Since the study presented here can only give a coarse

overview of the electron-microscopic zircon research, profiles of the three important elements Hf, Y, and U are

presented.

The zircon crystal separated from the peraluminous S-type rock (see also grain 1 in Fig. 5) represents a typical

example for the type of inhibited or transitional prism growth noted in the preceding section. Along the profile line A-B

marked in Fig. 7 growth of the prism {110} may be subdivided into two different stages: The first stage is characterized

by large growth increments of this prism, causing its continuous reduction in size. The second stage is marked by a

more or less undisturbed growth of {100}. From a crystal-chemical point of view, this remarkable change of prism sizes

in the innermost growth zones of the studied grain is caused by heavy elements, especially U and Hf, blocking the

growth of one prism form but having no influence on the growth of the, let’s say, competing prism form. In the concrete

case, this blocking effect becomes visible in the inner bright zone of the zircon crystal, where concentrations of U and

Hf are increased (see profile), leading to a growth blockade of the form {110} but an undisturbed growth of the form

{100}.



______________________________________________

A more undifferentiated picture is given for the zircon crystal from high-K calcalkaline rock, where due to the

undisturbed growth a ‘switch’ of prism forms during growth is not available. Prism growth, however, is marked by a

continuous enlargement of the form {100}, whereby bright and dark growth bands are arranged according to a regular

pattern. Regarding the chemical element profiles, Y behaves rather sensitive along the growth zone pattern in the way

that element concentration in bright zones is significantly increased and vice versa. The elements Hf and U do not

follow any similar trend. Summing up, undisturbed prism growth does not exhibit any specificities concerning element

insertion into the crystal lattice.

Fig. 7 Chemical profiles of the three elements Hf, Y, and U on selected cross sections of zircon crystals separated from a

peraluminous S-type rock (a) and a high-K calcalkaline I-type rock (b). Profile length corresponds to 20 µm, respectively.

5. Conclusions

As demonstrated by several examples, electron microprobe analysis of magmatic minerals including BSEI and WDS

measurement may be regarded as an unavoidable tool for the answering of important questions in earth sciences which

are not solved appropriately hitherto. While BSEI provides a reliable method for studying the growth development of

single crystals, chemical analysis using the WDS enables the deciphering of possible correlations between magma

chemistry and the course of crystallization.

In the study presented here electron-microscopic and micro-analytical techniques were applied to accessory zircon,

representing an ideal mineral for growth analysis due to its partly extensive formation of growth bands. For both

pyramidal growth, involving the forms {101} and {211}, and prismatic growth, involving the forms {100} and {110},

three different developmental trends, i.e., undisturbed growth, inhibited or transitional growth, and pathological growth

(Fig. 8), were found, whereby each growth type preferentially occurs in a certain rock type characterized by specific

concentrations of major and trace elements. Although the relationship between the mode of zircon growth and host rock

chemistry has been documented for several times in the past [4, 13-15], zircon crystallization is not completely

understood in all its different facets hitherto.

As a main progress in the scientific field of crystallography it could be found that growth of zircon crystal faces from

the magmatic source is among other influenced by insertion of certain trace elements (e.g., Hf, U, Th) into the crystal

lattice. Concerning the growth of the prismatic forms {100} and {110}, high concentrations of these elements have a

disturbing effect on the growth of {110}, whilst the growth of {100} does not seem to be inhibited, resulting in a

transitional growth phenomenon [13-15]. On the other hand, low concentrations of these elements have a somewhat

disturbing effect on the growth of {100}, whereas the growth of {110} remains unaffected. This circumstance results in

the “pathological” growth phenomenon introduced above.



______________________________________________

Fig. 8 Sketch summing up the three different types of prismatic and pyramidal growth that have been exhibited in this study with

the help of BSEI techniques.

Summing up it can be concluded that magma chemistry represents a highly essential environmental factor

influencing zircon growth significantly. By using appropriate electron-microscopic techniques, the value of zircon as a

petrogenetic indicator becomes clearly noticeable.

References

[1] Sunagawa I. Growth of crystals in nature. In: Sunagawa I, ed. Materials science of the Earth’s interior. Tokyo: Terra Scientific;

1984:63-15.

[2] Caruba R. Morphologie de zircons synthétiques: corrélations pétrogénétiques. Canadian Mineralogist. 1978;16:315-323.

[3] Caruba R, Baumer A, Hartmann P. Crystal growth of synthetic zircon round natural seeds. Journal of Crystal Growth.

1988;88:297-302.

[4] Vavra G. On the kinematics of zircon growth and its petrogenetic significance: a cathodoluminescence study. Contributions to

Mineralogy and Petrology. 1990;106:90-99.

[5] Vavra G. Systematics of internal zircon morphology in major Variscan granitoid types. Contributions to Mineralogy and

Petrology. 1994;117:331-344.

[6] Speer JA. Zircon. In: Ribbe PH, ed. Orthosilicates. Washington, DC: MSA;1982:67-112.

[7] Pupin JP, Turco G. Une typologie originale du zircon accessoire. Bulletin de Soieté Francais en Minéralogie et Cristallografie.

1972;95:348-359.

[8] Pupin JP. Zircon and granite petrology. Contributions to Mineralogy and Petrology. 1980;73:207-220.

[9] Poldervaart A. Zircon in rocks: 2. Igneous rocks. American Journal of Science. 1956;254:521-554.

[10] Frasl G. Die mikroskopische Untersuchung der akzessorischen Zirkone als eine Routinearbeit des Kristallingeologen. Jahrbuch

der Geologischen Bundesanstalt. 1963;106:405-428.

[11] Paterson BA, Stephens WE, Herd DA. Zoning in granitoid accessory minerals as revealed by backscattered electron imagery.

Mineralogical Magazine. 1989;53:55-61.

[12] Paterson BA, Stephens WE, Rogers G, Williams IS, Hinton RW, Herd DA. The nature of zircon inheritance in 2 granite plutons.

Transactions of the Royal Society of Edinburgh: Earth Sciences. 1992;83:459-471.

[13] Benisek A, Finger F. Factors controlling the development of prism faces in granite zircons: a microprobe study. Contributions to

Mineralogy and Petrology. 1993;114:441-451.

[14] Sturm R. Longitudinal and cross section of zircon : a new method for the investigation of morphological evolutional trends.

Schweizerische Mineralogisch Petrographische Mitteilungen. 1999;79:309-316.

[15] Sturm R. Imaging of growth banding of minerals using 2-stage sectioning: application to accessory zircon. Micron.

2004;35:681-684.

[16] Vavra G. A guide to quantitative morphology of accessory zircon. Chemical Geology. 1993;110:15-28.



___________________________________________________

analyzing growth kinetics of magmatic crystals by backscattered

Documents