Central Laboratory of Mineralogy and Crystallography, Bulg. Acad. Sci., Sofia, Bulgaria

 Introduction

As early as in the XVII century N. Stenonis drew attention to a peculiarity in the growth of quartz crystals. According to him the new substance of the crystal does not join equally all faces but mainly certain faces on the top which he calls end faces, i. e., the faces of the small z{01.1} and the big r{10.1} rhombohedra. Unlike the rhombohedra which have layers of growth, N. Stenonis has distinguished the hexagonal prism m{10.0} as one that lacks such layers and increases its area on the account of the rhombohedra’s growth. Even till now this statement is not accepted synonymously by all investigators. Main reason for this is that the most characteristic habitus of the natural quartz crystals is that of the short prism. The ratio of the crystal thickness (along the two-fold axis) to its height (along the three-fold axis) in it varies from 2:3 to 1:4. Only a few finds of quartz crystals with ratio 1:10 have been published. If we assume that the growth of quartz is mostly due to the rhombohedra z{01.1} and r{10.1} than the mentioned ratio should be at least 1:100.

With the development of the piezoengineering and the receiving of huge amounts of hydrothemally grown man-made quartz it has become clear that the perfect crystals do not grow in direction of the hexagonal prism m{10.0} or in the best case their growth rate does not exceed 0.00n millimeters per day.

In this case if we accept Stenonis hypothesis we can not explain the observed morphology of the natural quartz crystals and if we reject it we will be in contradiction with most of the experimental data.

This work is an attempt to explain one of the possible mechanisms for growing of the hexagonal prism. Considerations about the natural quartz crystals habitus are also presented here.
 

Experimental procedure

All runs were carried out in two-zone tube furnaces. Unlined steel autoclaves of 150 ml were used. The filling was 85% to avoid formation of heavy cakes. Crystallization proceeded at 370 ±2.4 °C at a temperature difference of 30 ± 4.8 °C. Temperatures were measured with chromel-alumel thermocouples fixed in holes bored in the autoclave outer wall. Strongly rounded pebbles of natural quartz gravel (milky quartz) with diameters of about 8 mm were used as a nutrient. The solvent used was 0.5 M NaOH solution with some LiNO₂ added. Three different types of seeds were prepared for the runs: man-made quartz Z-cut plates; pure natural quartz crystals, elongated along the c axis and three pieces cut from a single elongated along the c axis natural quartz crystal and fixed together with a wire.

The grown crystals and their cuts were investigated by means of microscopic reflected and transmitted light. Special attention was paid to the morphology of the grown faces and to the presence of Brewster’s figures. Sections cut from all crystals were subjected to weak superficial dissolution in NH₄HF₂ to reveal the zone and sector structure.
 

Results

As seen from Table 1 growth of the hexagonal prism has been observed only in those runs in which natural quartz crystals have been used as a nutrient. In the process of growing the three fixed together natural quartz crystals formed a single crystal without visible defects on the hexagonal prism (Fig. 1).

The morphology of the grown faces of the hexagonal prism owns some specific features. No horizontal striation has been observed macroscopically. The surface is flat with some small needle-shaped quartz crystals coming out of it. In reflected light it can be seen that the whole hexagonal prism is covered by tiny four-, five-, or six-lateral step-shaped hills (Fig. 2). Their faces are not equally developed. The hills have one almost vertical short face parallel to the main crystallographic axis and surrounded by three or four intensively stepped faces. The steps are not always straight. Often among three adjacent faces only the steps of the middle one are arched while on the other faces these are comparatively straight (Fig. 2). The height of the hills reaches 1-5 mm and the area they occupy is about 10-30 mm.

Typical Brewster’s figures have been observed on polished pinacoidal cuts in transmitted light and crossed nicols in those sectors where growth of the hexagonal prism takes place. This is a sign for twinning after the Brazil twin law. The amount of Brazil twins is considerably bigger at the entrance angles where the growth rate of m{10.0} reaches almost 0.1 mm per day.

Fig. 4 clearly presents layers of dissolution which mark the growth stages of the crystal. Two systems of mutually crossing dissolution layers can also be seen on Fig. 5 and their orientation explains their nature. No doubt the vertical layers mark the growth stages of the hexagonal prism (see also Fig. 4). The crossing them finer layers have a rhombohedral orientation and their appearance is discussed in the following part.

Discussion

The presented results are indicative of a couple of different mechanisms for the quartz crystals growth. On one hand in the presence of Brazil twins and dislocations high growth rate of the hexagonal prism compatible with that of the main rhombohedrons can be detected. This suggestion is supported by the observations made on natural amethyst and citrine crystals. Most of them do not have strongly expressed horizontal striation on the faces of their hexagonal prisms. The thickness to height ratio in their crystals is of the order of 1:1 and all they are Brazil twins. In the model of the Brazil twinning boundary in the structure of a-quartz the transition angle Si-O-Si equals 138.5° and it is always 5.5° less than the ideal value of the same angle in untwinned crystals, i.e., 144°. This non-essential difference is most probably removed by incorporation of isomorphic admixtures of [Al³⁺]Li⁺ or [Fe³⁺]Li⁺, what is considered to give the radiation citrine and amethyst coloring. Incorporation of admixtures brings to elevated concentration of dislocations in these sections. The growth layers, generated by them form hills which face morphology is well presented on Fig. 2. These hills will never close in the process of growing, thus becoming a constant source for new growth steps. As a result the growth of the hexagonal prism is realized.

On the other hand one can not doubt that in some cases the quartz monocrystals grow mainly on the account of the rhombohedral faces. This fact is supported by numerous observations on colorless quartz crystals which are more elongated along the c-axis than the typical amethysts.

The nature of the presented on Fig. 5 system of layers can be explained as follows. The vertical layers (parallel to the hexagonal prism) mark the growth stages and bear information about the history and the characteristics of the crystallization process. The layers with the rhombohedral orientation are manifestation of the crystal structure peculiarities and the direction of the strong linkages (PBC vectors) in quartz. These latter layers can interlace with the real growth layers thus covering them and marking an untrue zone or sector structure of the crystal.

Conclusions

1. In the presence of Brazil twins or a big amount of dislocations the growth rate of the hexagonal prism m{10.10} increases and comes closer to that one of the main rhombohedra. Growth pyramids with specific morphology are formed in the process of m{10.0} growing.

2. When Brazil twins or dislocations are missing the quartz crystal growth is realized mainly by the rhombohedra z{01.1} and r{10.1}. In this case the typical for quartz horizontal striation appears.

3. Integration of three or more crystal specimens is possible with a subsequent formation of a single but already twinned quartz crystal.