1.1.6 Grading Glass and Graded Seals

metal so thin, it becomes mechanically weak. Another option is to use Kovar® (an alloy composed of cobalt, iron, and nickel), which has a relatively low thermal coefficient of expansion. Although it requires two grading glasses to seal it to common laboratory borosilicate glass, Kovar does not have to be machined thin and therefore maintains greater strength than do machined metals. Often the advantages far outweigh the extra effort involved in making the graded seal. Kovar is easily soldered or welded to other metals for vacuum-tight seals.

1.1.7 Separating Glass by Type

Because laboratory glassware may be manufactured from a variety of different types of glass, such varied glass can become mixed up—potentially leading to confusion or later problems. It is important to either maintain different types of glassware separately or be able to tell them apart. Although the former approach is preferred, the ability to identify and separate glass is important not only to save time, but also for safety and even the integrity of your experimentation.

Typically, only a few commercial soft glass items may work their way into a research lab. Such items as student ware graduated cylinders or burettes are readily identifiable, and since these items shouldn't be heated, are unlikely to cause damage. Specialized or custom made glassware may not be as easily identifiable and therefore may require some analysis.

Soft glasses, hard glasses, and high temperature glasses all look the same—like clear glass—and therefore may be hard to separate. The process to separate glass may be either destructive or nondestructive. Destructive techniques involve doing some permanent physical change to the glass, after it which will never be the same. Nondestructive techniques, obviously, are preferred if you wish to preserve all of your laboratory glassware. A listing of important properties of various glasses are shown in Table 1.2.

Logical Deduction and Observation (Nondestructive). By logically deducing what you are likely to have in your laboratory, it is often simple to separate different types of glass and glass apparatus. For example, fused silica or quartz glassware is expensive. Unless there is a specific demand or need for fused silica in your lab, it is extremely unlikely that you have any. Thus, there is generally no need to look for things that are not likely to be there. (One exception to this is if you have any UV cells.)

Glassware identified with ceramic decals or raised letters saying "Pyrex," "Kimex," or "Duran" are exactly what they say they are.* Pipettes from Kimble may be identified as "Kimex-51®" and are not the same as Kimble's common laboratory borosilicate glass (KG 33). Pipettes from Corning use a lab glass called "Corex®," which is also different from Coming's common laboratory glass (7740). If a commercial container (as opposed to something custom made)

Prolonged cleaning by base baths or HF can remove ceramic markings.

Table 1.2 Characteristics of Specific Glass Types"

" Except where otherwise noted, these data are compiled from Properties of Coming's Glass and Glass Ceramic Families by Corning Glass Works, Corning, New York 14831, ©1979. Ranges indicate a small family of glass types with similar characteristics.

* This glass is for radiation shielding. The data are from Engineering with Glass by Corning Glass Works, Corning, New York 14831, ©1963.

c This type of tubing is commonly used on neon signs.

rfThis information is based on Kimble Glass type R-6, compiled from Kimble Glass Technical Data by Owens-Illinois Inc., Toledo, OH 43666, ©1960.

' Most laboratory glassware is made from this glass.

^Pipettes and other pharmaceutical items are made from this glass.

gAluminosilicates are used for ignition tubes, for halogen lamps, and for containing helium.

hVycor® is 96% silica glass.

' Fused silica is essentially pure silica with few impurities.

doesn't have the words Pyrex or Kimex somehow attached to it, it is invariably made of soft glass.

Glassware that has bends and/or has been fused to other pieces of glass will be either common laboratory borosilicate glass or fused silica. Any soft glass made into laboratory apparatus is for special application or, more likely is very old (and ought to be in a museum).

Sighting Down the End of a Glass Tube or Rod (Nondestructive). Even though most glasses will look the same when observed from the side, looking at

glass end-on will exaggerate the different colors inherent within the glass. Common laboratory borosilicate glass will typically show a soft pale blue or green shade of color. Soft glass (soda-lime) will typically exhibit a fuller deep blue. It is recommended that you obtain a 15to 20-cm rod of each type of glass to keep on hand for use as comparison samples. Absolutely clear (water white) glass is hightemperature fused silica or Vycor. By shining a deep UV (254 nm) lamp on the side of the glass and sighting down the end of the tube, Vycor will display a yellow/green or colorless fluorescence while fused quartz will be strong blue (do not stare directly at the UV light).

Matching the Index of Refraction (Nondestructive). To match the index of refraction for two or more glasses, obtain a liquid with the same index of refraction as one type of glass. When an unknown glass is put into the liquid, the glass will seem to disappear if it has the identical index of refraction. Two standard solutions for identifying Pyrex or Kimex glass (their refractive index is 1.474) are:

1)16 parts methyl alcohol

84 parts benzene

and,

2)59 parts carbon tetrachloride

41 parts benzene.

The major problem with both of these solutions is their toxicity. The first solution should only be used in a fume hood, and the second is so toxic that it should not be used at all. An alternative solution, albeit a bit messier, is common kitchen corn oil. It does not match the index of refraction as closely as do the two solutions mentioned above, but it does the job and is safe to use. The corn oil can be removed with soap and water.

Checking the Sodium Content of the Glass (Semidestructive). Take the bottom unglazed side of a white ceramic tile and evaporate onto it a small quantity of a phenolphthalein-ethanol solution. To test a piece of glassware, drip a little water onto the plate and draw a line across it with the glassware. The action will lightly scratch the surface, so choose a nonsignificant spot. By scratching the surface, the alkalis of a soda-lime glass will be exposed to the phenolphthalein solution and a pink line will form across the plate. Borosilicate glass may exhibit a minor color

Fig. 1.3 Step process for using linear expansion to separate glass types.

Bottom right section of an

Erlenmeyer flask

Surface (fire) checks

Fig. 1.4 Surface (fire) checks caused by placing hot glass on a cold surface.

(fails). On the other hand, metals (even spring metals) can remember a new shape or position if flexed beyond a certain point. This quality is true for tangential as well as axial rotations. However, unlike metals, glass gives no indication that it is about to break, nor does glass provide any indication of where it is receiving stress* (for example, glass does not begin to fold prior to fracture).

Once flexed beyond a given point, glass breaks. The amount of flexure required to achieve that given point is dependent the nature of the flaws in the region being flexed. Glass can only be broken if two conditions are present: flaw and stress (more specifically, the stress of tension, not compression). Because glass is perfectly elastic until the point of failure, the location of the failure will occur at the most susceptible flaw. This property is a sort of weak-link principle: Glass under tension will break at the weakest link, which, by definition, is its most vulnerable flaw.

Tests demonstrate22 that a one-quarter-inch-thick piece of glass that has received normal handling can withstand pressures of 6000 psi. Sandblasting the surface to provide large numbers of flaws drops the potential strength to some 2000 psi. However, if a fresh surface is acid-polished and then coated with lacquer (to prevent further abrasion by handling and to limit water content), strengths of up to 250,000 psi have been achieved.

Despite its tremendous strength, glass can be fairly easy to abrade. On the Mohs' scale of hardness,* glass is between #5 (apatite) and #7 (quartz). So, glass can scratch materials with lower numbers (for example, copper, aluminum, and talc). Likewise, glass can be scratched by materials that have higher numbers (for example, sand, hard steel, and diamond).

We know that diamond can scratch glass, but so can hard metals such as the hardened steel of a file or a tungsten-carbide glass knife. A beaker that is slid or

*An exception to this statement would be viewing glass through polarized light, which provides excellent visualization of strains within glass. However, this requires special equipment (polariscope), and the glassware must be in a particular orientation with respect to the equipment.

fA ten-scale division of geologic minerals used by geologists to help type rocks, with talc as #1 and diamond as #10.

Note that K, A, and B are dimensionless. The only criterion is that A and B be in the same measurement units. Some sample calculations using Eq. (1.1) can be seen in Table 1.3.

Glass can be broken at a specified location by creating a flaw where the break is desired and then applying stress (tension) to that flaw. However, an improperly made flaw can easily result in a flaw of undesirable quality. For example, an inexperienced flat glass cutter will take a wheeled glass knife, aggressively bear down on the glass, and push the cutter back and forth several times. While a single scratch could have broken cleanly and easily, the repeated scratches create a heavy, round bottom groove with many side fissures. The resulting crack is likely to drift off to the side and not follow the heavily gouged crevasse. Alternatively, an inexperienced glass rod and tube cutter will take a triangular file and (again) bear down on the glass and saw the file back and forth for a "real good scratch." Unfortunately, the nicely rounded fissure will also not break easily. The extra force required to break the heavily gouged flaw is likely to cause the glass to break into many fragments, creating a serious risk of injury.

The key to successful glass breaking is a thin, deep scratch. However, you do not achieve effective depth by excessive force or repeated scratching. To cut a soda-lime flat glass pane, you must maintain a firm pressure on a rotating wheel cutter and make a single continuous scratch toward you. Using a yardstick or meter stick as a straight line guide is recommended. If you bear down extra hard on any part of the swipe, you are likely to create side fissures that may cause the crack to sweep off to one side. The same technique applies to borosilicate flat glass, although more pressure will have to be exerted to achieve the same depth of flaw.Regardless, it is important to maintain even pressure throughout and to make only a single stroke.

Although not as effective as a tungsten-carbide glass knife blade, a triangular file can be used on glass tubing. The trick is to make a single, fast swipe of the file toward you and not to saw back and forth. In addition, the sharper the file's edge, the narrower the scratch. One way to ensure a sharp edge on the file is to use a grinding wheel to remove one face of the file (see Fig. 1.5). During the grinding process, constantly lower the file into a water bucket to cool the metal. If the file gets too hot, it will turn a bluish color, indicating that the file has lost its temper and will dull faster.

It has been well demonstrated since the 1940s that glass breaks more easily if wet with water than if dry. Dry glass is stronger than wet glass (by 20%), and

I K\\\\) I

Grind a face from one side of the file to obtain two sharp edges

Fig. 1.5 Altering a triangular file for more efficient glass cutting.