A Closer Look at the Crystalline Structures of Snowflakes and other common objects

Optics 307/407: Scanning Electron Microscopy (May 2005)

Author: Jeff Bentley, jb007k@mail.rochester.edu

Our environments constantly surround us with many miniscule and intricate structures. Crystalline structures, being one of the more fascinating types, come from the straight from the earth, are synthetically molded, artificially processed, and even occasionally just fall from the sky. As will soon be displayed, crystalline structures retain their shape due to a complex series of linear and geometrical structures. In this Project, we will observe four different types of common crystalline structures; amethyst rock, commercial glass, salt, and snowflakes.

The samples have been prepared by the Gold Sputter Coating technique and a very unique method of Snowflake Collection. They will be examined using the Secondary Electron technique, an Electron "Flight Simulator" Projection program, Multiple Colorization techniques, X-ray EDAX "SEM Quant ZAF" Compositional Analysis, and Stereo Pair Imaging. Overall, this project is intended to supply a general overview of similarities in crystalline structure and will to also carve out some useful ways to apply the Colorization process to images in an effort to expdite and enhance viewer comprehension.

Sample 1

Crystals, a form of rock, can be found in the Earth in a naturally crystalline form. This form gives them their luster and unique shape, while providing them with varying hardness'. Using the data from the SEM Quant ZAF process' Compositional Analysis, we see that the magenta and purple Amethyst crystals contain mostly Silicon Oxide (SiO2). We enter this chemical composition into the Electron "Flight Simulator" Program and find that our Secondary Electron will penetrate very nicely into the material.



In preparation for viewing, this sample was securely held in a vice-lock, while some smaller pieces were chipped off of it. These pieces came off in releatively long shards and were still very resistant to shatter. The pieces were sputter coated for 60 seconds at 20mA of current, but an additional 20 seconds at 20mA was necessary to reduce excessive charging. Below, a lower magnification (right hand) image is compared to a higher magnification image (left hand).

Below is the first instance of colorization with the purpose to increase the ease with which viewers can process and understand data. Often times, viewers can be confused image magnification and will draw false conclusions, but not anymore. A "Magnifiation Bar" has been setup based on a spectrum of color changing from magenta to dark purple. The colors on the bar are very similar to those found on an actual Amethyst crystal, and represent increases in magnification as they darken. This colorization technique allows the viewer to easily distinguish multiple images of different magnification, and easily identify the sample they are viewing.
Notice that as the magnification increases, we see redundant characteristics in the amethyst crystal as they form long solid geometrical bands. We can thus hypothesize that these long geometrical bands are what give the crystal its rigidity and extreme support, while also making it sharp to the touch. All of the smaller particles on the samples are most likely not from the rock, since they do not seem to match surrounding structures. They are most likely due to dust, mold, or dirt in the structure of the rock crystal.

As a further aid to the viewer, and to achieve a greater feeling for the crystalline structure's orientation, we see the figure below is a Stereo Pair Image. Stereo Pair Images are a series of two images whose only difference is the 4 to 8 degree tilt of the stage at which they were imaged. They provide a three dimensional view of the region under inspection, and in this case can offer the viewer a much greater understanding of the linear and angular composition. Stereo Pairs are generally viewed with the red panel over the left eye and the blue panel over the right eye.


Sample 2
Commercial Glass

Commercial glass is the next sample observed for its crystalline structure. This particular sample consists of glass fragments from a Snapple bottle, something which is generally discarded and rarely ever viewed with interest. From our Compositional Analysis we see that the glass consists of mainly Silicon, Oxygen, and a small part Sodium. This fits perfectly with the Soda-lime type of glass which consisting of, 71-75% SiO2, 12-16% NaO, and 10-15% lime (1). Fortunately, these types of containers are generally inert, thus they are frequently used in the commercial industry as jars, bottles, and other glass object which can contain safely ingestible items (i.e. food, drink, etc...). This sample was, similar to the Amethyst case, gold sputter coated once at 20mA current for 50s, and followed up by an additional 20s at 20mA to reduce charging.

Once again, our Electron "Flight Simulator" Program shows that the sample will be very receptive to Secondary Electron analysis. The above electron projection only uses 3300 electron trajectories, any greater amount begins to completely fill in the chart.
Below, we see a low magnification (upper left), medium magnification (upper right), and a higher magnification (bottom center) micrograph of the glass under examination. Very similar to the amethsyt crystal, the soda-lime glass breaks off in long angluar bands. However, a major difference is the length and thickness of these bands in comparison to the amethyst; the glass' seem shorter and thinner. Since glass tends to shatter easier than rock crystal, the length and thickness of these bands seem to have an impact on the durability and hardness of a crystalline structure. The difference in length is especially noticeable in the glass fragments where shorter bands prevail, and allow smaller cube-like particles to form.

Once again, the same colorization technique is used with the glass sample. Green was chosen to help the viewer relate the images to the soda-lime glass type (lime being associated with green) as opposed to confusing it with the amethyst images. These series of magnifications show similar striations to the earlier amethsyt case, yet have shorter bands which seem more easily broken. Overall, this supports the growing hypothesis that crystal band length and thickness are proportional to the durability and hardness of the crystalline structure.

The Stereo Pair Image for the glass shows a unique looking break point in an effort to illustrate how the sample breaks on the bands, and has a very angular structure to it.


Sample 3
Everday Grains of Salt

Although we eat it nearly every single day, salt's amazing internal structure is frequently overlooked. The salt grains under examination were plucked from the sides of Rold Gold Mini Twist pretzels. After "harvesting" the grains, they were mounted via carbon tape onto a silver platform stud and mounted into the sputter coating device. An initial coating using 20mA for 50s resulted in excessive charging, thus the samples required an additional 20s at the same current. Our "SEM Quant ZAF" Compositional Analysis confirms the fact that we do indeed have NaCl, a.k.a. table salt. From here, the NaCl projection, below, shows the expected electron scattering within the samples.

This projection only observes 3300 trajectories, any more than that will almost completely fill the chart. Therefore, our salt will definitely produce nice images when we use Secondary Electrons. Below, our first image (left) shows a general low magnification view of a single salt grain, while the second image (right) shows a magnified view. In contrast to the amethsyt and soda-lime glass samples, the crystalline structure of salt seems to be very cube like and essentially has very short bands with relatively small thickness. This supports the hypothesis that short and small bands have weaker structures than long and thick ones, since salt is an easy material to break.

Our colorization continues, as the salt images are placed on a color/magnification scale from yellow to red. These colors were mainly chosen to differentiate the salt from the other two samples, since salt has nearly no correlation to any of the displayed colors. These images continue to support the hardness/band size hypothesis since their cubic shape becomes more and more prevalent as we examine them closer.

These two Stereo Pair Images very nicely show the cubic composition of salt, and more importantly show how the structures overlay each other as they progres sthrough the material.


Sample 4
Snow Flakes

Now that we have examined some of the more obvious crystalline strucutres present in our everyday life, it's time to look at something a little different. Snowflakes also have a crystalline structure, and are just as natural as any of the previous samples. The method for snow flake catching involves using a 1% or 2% Formvar in Chloroform solution to make a mold of the snowflakes, for examination. A stud is droppered with the solution (both are pre-cooled to match environmental conditions) and snowflakes are caught on a close knit cloth (a dense fleece scarf was used in this case) then transferred via solution-dipped toothpick to the stud. When left outside, the cholorform and water sublimate, leaving a formvar mold of the snowflake (2).
Unfortunately, this process is not perfect and frequently leaves formvar residue over the vacant snowflake structure. To deal with this issue, we use colorization to once again expedite and enhance viewer understanding of the sample. This is done by outlining the "theorized" structure of the snowflake under examination. After closely examining the image, a colored snowflake structure is placed over the image, based on structural cues, which outlines the actual structure of the snowflake. The colorization process gives viewers guideline by which to clearly see the snowflake structure upclose. In our case, the snowflake structure is consistant with the amethsyt, glass, and salt, in that snowflakes rely on a geometrical structure to maintain their shape.

The above images suggest that snowflakes use a hexagonal structure, as opposed to the more earthly based samples' cubic structure, to maintain their rigidity and resistance to air and low atmospheric pressure. Below, there are Stereo Pair Images which more properly display the angular and geometrical structure of the snow flakes. Unlike the other Stereo Pairs, some of these ones are best viewed with the red panel over the right eye and blue over the left eye.



Overall, the Amethyst, Soda-lime Glass, and Salt crystalline structures show the constant trend of cubic type bands which vary in length and thickness. This project shows evidence, via imaging, that crystalline structural hardness increases when these bands are long and thick, and decreases when the bands are short with a small thickness. This is proven correct when we observe that Amethyst rates a 7 on Moh's Scale of Hardness (3), Soda-Lime Glass rates thinner score of 6-7 (4), and Salt with a 2.5 (5). These results are more efficiently shown using the above Colorization techniques. The snowflakes share the samples' angular structure which is the base for crystalline strength. They do not rate on the typical hardness scale, but they do resist air pressure when forming and falling to the ground. Thus, the angular crystalline structure seems to fortify snowflakes, as well as the Amethyst, Soda-Lime Glass, and Salt, to allow them greater beauty and durability. It should also be noted that I attempted to make a formvar mold of freezer produced snowflakes, but they came out very poorly since they didn't have the atmospheric conditions necessary for strong crystal formation.



  1. "Types of Glass." Glasstopia. May 2005 <http://www.glasstopia.com/e_site/glassis/category/species/bychemical.html>
  2. The snowflake method was supplied by Brian McIntyre
  3. "Physical Properties of Amethyst for Identification and Classification Purposes." JewelrySupplier.com. May 2005 <http://www.jewelrysupplier.com/2_amethyst/amethyst_properties.htm>
  4. "Soda Lime Flat Float Glass." Valley Design Corp. May 2005 <http://www.valleydesign.com/sodalime.htm>
  5. "What is Salt?" Salt Institute. May 2005. <http://www.saltinstitute.org/15.html>




Please enter any comments, criticisms, questions, etc. below.

Your name:

Email address: