It has been said that the light microscope has been the most importantscientific instrument of all time. It is the tool of micromorphology. "Micromorpholgy is the branch of soil science that is concerned with the description, interpretation and, to an increasing extent, the measurement of components, features an dfabrics in soils at a mciroscopic level" (Bullock et al, 1985). Optical microscopy is one of the few techniques that allows us to examine the soil and its components in situ, unaltered and undesturbed by preparation or analytical procedures. Micromorpholgy usually implies thin scetions of the soil material; a 2-D plane view of a 3-D body.

the preparation of thin sections of soil and geological material is not new. The basic procedures have not changed for over 100 years. What has changed is that we now have support facilities, equipment, tools, and supplies specifically designed to help us produce high quality sections of diverse materials. More recently, we have also developed a collection of concepts and terms that aid us in describing an dinterpreting our observations, and conveying the concepts to others. Likewise, computerized image analysis is being developed which will enable us to quantify many features and tackle new problems. These new advances in tools and prepartion techniques have opened up micromorphology to a host of scientific disciplines including pedology, mineralogy, geology, engineering, materials science, soil physics and chemistry, environmental engineering, land use and management, etc.

In the past, most thin scetion preparation was manual, tedious, and labor intensive. In that period, thin section preparation was more an art than a science. Today, machines are available (often at considerable cost) which remove much of the guess work an dputs preparation on an automatic or "assembly line" basis. Despite advances in technology, the critical steps of sampling, drying, and impregnation are vital for correct interpretations. Much of the material outlined here is discussed and illustrated in more detail in Murphy (1986).

Sampling is one of the most critical steps in micromorphology. Errors or hasty decisions made here may affect what you see in thin sections and your ultimate interpretations and conclusions. It is critical that smaples obtained for tin section be representative of the soil or material as it occurs in nature. The sample must also represent an unbiased selection. A representative sample is important when you consider the volume relationship between an thin setion, the clod sampled, and the extent (lateral and vertical) of the horizon or material being sampled. A typical thin section may be <0.05 cm³ whereas the impregnated block from which is came may be 500 cm³, and difference of 10,000. Because sampling is critical, it does not begin in the field, but in the office. There are several aspects of sampling that should be considered before going to the field:

1. Purpose of the Investigation (Why take a sample?)
2. Site Selection and Supplemental Data form the Site
3. Sampling the Soil (Size, Orientation, Location)
4. Timing of the Sampling (Temporal Variability)

1. Purpose of the Investigation:
   There must be some overriding purpose for obtaining a sample for micromorphology. The purpose of the investigation may dictate the size and the number of samples per horizon, sample orientation, the number of horizons or subsamples collected, and possibly the sampling techniques employed. The purpose may influence how the samples are obtained; from hand dug shallow pits or large and deep trenches dug with a backhoe. Is micromorphology the primary purpose of the study, or will micromorphology simply supplement information for a larger more comprehensive study? Are the micromorphological data to be quantitative, qualitative, or descriptive? Such questions need to be answered before one reaches the field. Each purpose is different as is each soil, so no absolute criteria can be presented.

2. Site Selection and Supplemental Data:
   As in any scientific investigation, the results can only represent the material sampled, not always adjacent fields, test plots or similar material. The site must be selected to be representative and meet the purposes of the investigation. If one sample is selected, or a complete profile, the site should be fully described following standard procedures. This includes site location, landscape position, and other information used to describe soil profiles and horizons. Bulk samples should also be taken for supplemental laboratory analysis. When describing a thin section, the scientist should have available a complete profile and horizon description. Knowledge of teh geological setting, topographic postion, and other pertinent information will aid in the interpretation. If there are descrepancies between the section and the profile description, something is amiss and needs to be investigated.

   NOTE Under no circumstances should "grab samples" be collected for micromorphological investigations. There are permitted if the purpose is to obtain a teaching sample that may have a unique meral assemblage or show some unique feature. Even so, a site location and brief description is useful.

3. Sampling the Site:
   In sampling a soil for micromorpholgy, several factors should be considered. The size of the ample collected in the field depends on the purpose of the investigation, the size of the features being investigated, the number of thin sections to be cut from the ample, the size of the thin section. Most laboratories can handle a microscope slide of 25x75mm (1x3") or 50x75mm (2x3"). In such cases, several sections can be cut from a sample with a volume of 300-500cm³. Some features such as illuvial argillans, root or animal infillings, slickensides, or tillage effects are not uniformly distributed, at least at the scale of a microscope slide. Other features occur at intervals greater than the dimension of the microscope slide and may be missed entirely or observed in concentrations greater than their true distribution. In these situations, multiple sections may need to be prepared form a large block, or the material may need to be subsampled, and sections cut from each subsample. Larger thin sections up to 100 by 200mm (4x8") have been prepared, but they require special equipment generally not available in most laboratories. Because most soils are friable, they need to be impregnated with a resin. the size oth the impregnating chamber or cost of resin may limit the size of the sample collected.

   Soil is not a homogeneous mix, but may have features exhibiting preferential orientation. Translocation of constituents usually occurs vertically, but horizontal bedding may also be present, and their expression will be dependent of the cut section in reference to the feature. The orentation of certain features may yield improtant information of clay illuviation, formation of secondary products, depostion processes, water infiltration, etc. The vertical orientation should always be marked on the sample and the sample container. When possible, both vertical and horizontal sections should be prepared form the sample unless the material is highly stratified or prior investigations yield information on feature orientation.

   If a complete profile is being sample, then samples should be obtained form teh center of each horizon. Horizon boundaries should be avoided unless there is a unique boundary condition. In some situations, such as comparing adjacent tillage plots, or plowed and fallow fields, it may be desirable to sample at reular depth intervals, especially where natural horizons have been altered.

   How many replicate samples does it take to represent a heterogeneous soil material? This question has been asked repeatedly, and a few have attempted to answer the question (see supplemental reference list for articles on sampling). Part of the problem is that each soil is different, as are the research objectives. Replicaates, or size of the sample obtained, depend on the size of the feature investigated, and its spatial distribution. However, the number of replicates is generally governed by time and resources (money) availabe. If micromorphology is not a major thrust of the investigation, it is probably a waste of time to strive for statistical accuracy. For much of our research, tow clods are collected. One impregnated, and one keptin reserve, or impregnated with a different resin. Two or more thin section (horizontal and vertical) are then prepared.

4. Timing of Sampling:
   Many soil features vay temporally. Such changes may be due to shrinkage and swelling, or freeze-thaw in response to changing moisture or climatic conditins, structural changes in surface horizons due to tillage operations or natural changes through the cropping season, mineral dissolution or precipitation in response to changin soil climate. We must remember that most soils are not static, but dynamic, responding to outside perturbations such as moisture, temperture, plant and animals, and man's influence. Sampling a soil when it is too wet or too dry can cause problems. Wet soil may deform easily while dry soils may easily fracture. In both situations artifacts may form. Kikewise, the fabric and pore pattern may change seasonally in response to changing moisture or tillage practices. Thus, knowing the time of sampling and the associated responses is important when making interpretations.
   SAMPLING for MICROMORPHOLOGY

   As mentioned earlier, the sample collected should be representative, undisturbed, and oriented. there are several ways of sampling such that these conditions are met. The sampling technique used may depend on the size of the sample and structural integrity of the soil material. Most soils, except for loose sands, have enough clay or other cementing material to allow them to be sampled and transported to the lab with miminal fabric alteration.

   Friable soils can be sampled by various means. An easy means is to extract a clod about 500 cm³ and trim it to fit a pint paper container or bulk desity box. The orientation can be marked on the clod, or a staple or thumbtack inserted in the top to denote orientation. The clod is carefully wrapped in Al foil for protection and maintenance of moisture. the Al foil wrapper should also be labeled with the orientation, profile number, and depth. The wrapped clod is placed in a marked pint paper container or partitioned box for transportation to the lab. Do NOT write the identification on the container lid, as lids fall off and can be easily misplaced. The sampling method is fairly easy, but has disadvantages, especially in highly structures soils. In these soils the clod is often extracted along planes of ewakness. biasing the sampling, or it may not reain intact upon drying. However, in most situations, it will be sufficiently large so that several then sections can be prepared.

   Another tecnique is to use "Kubiena tins", disposable metal boxes made from aluminum, stainless stell, or galvanized sheet metal. They should be inexpensive and disposable so that the samples can be dried and impregnated in the boxes which maintains sample integrity decreases the chance of disturbing the sample. these are simple boxes with removable lids that fold over the sides. The side is made of one piece, but not soldered at one corner. The sides can be held in place with tape during sampling. Murphy (1986) describes these boxes. They can be fabricated to any size at a local sheet metal shop. The size of the box will be determined by the size of the thin section. We have used boxes as small as 50x100 by 37mm deep (2x4 by 1.5" deep) up to 90x165mm by 50mm deep (3.5x6.5" by 2" deep). the larger the sample, the more problems you may encounter with drying and impregation. One consideration when selecting a size is the availability of a container to hold the sample during impregnation. Disposable loaf pans are available in several sizes at most grocery and super markets at a modest cost. These can be purchased, and the Kubiena tins fabricated to slightly smaller dimensions so that the impregnating solution will cover the Kubiena tin. If you use some other size of Kubiena tin, you will have to fabricate your own container to hold the sample during impregnation. these can be made from Al flashing using wood former. If you make your own, check for leaks befor impregnating the sample.

   A modification of this technique is to use a metal moisture can, 80mm (3") in diameter and 50mm (2") tall. If a pint paper container is lined with Al foil, it will hold the moisture can during impregnation. The moisture can is used in very friable or loose sands. Holes are first punched in the bottom, the can inserted in the soil, and orientation marked on the can. The excess soil is trimmed flush with the surface and the lid put on. This will keep the sample intact. With holes in the bottom of the can, the sample can be impregnated in the can, thus avoiding any disturbance during removal.

   For both of these metal boxes, a smooth, flat face is first prpared at the disired depth and orientation, and the same size as the container. The soil is then trimmed back into the profile so the box will just slip over the exposed block. Roots are carefully cut with a knife or scissors. When the box is full, the box is carefully removed from the pit face by backcutting, ensuring an excess of material prjecting from the bax. The excess soil is trimmed off so the lid will fit. The box is then marked (on the side) with the sample number and orientation and sealed with tape. Even in very loose sands, the Kubiena tin may not easily be pushed into the soil; the sides may tend to bulge out makin the lid difficult to put on. Moisture cans, however may be pushed into some soils without distortion.

   For highly indurated soils, the Kubiena tins are more difficult to use. In most cases, cutting or chiseling out pieces can extract the soil material. A gas powered rock say may be useful in highly indurated materials such as petrocalcic horizons. these samples should be wrapped in Al foil for protection against moisture loss and disturbance during transport.

   The best policy in sampling is to use common sense. Take an adequate number of samples of sufficient size. It is easier to collect more samples than needed than to resample. Protect the samples during transportation, and be sure the samples are well marked with site designation, horizon, depth, and orientation. Read some of th ereferences listed elsewhere to obtain an idea of how other approached sampling.

Once the sample is collected in the field it is out of its natural environment, and susceptible to changes which can alter the fabric. This is especially true of soils that have a high clay content, or show evidence of shrink-swell. Highly structured soils may tend to fracture when dried. Theus, drying techniques are important to preserve the natural fabric. The sample needs to be completely dry because most impregnating resins are hydrphobic and will not mix with water. CIBA-GEIGY does make an Araldite resin that can mix with a small amount of water, but we have not tried that product. Wet highly organic soils cannot be dried without alter their fabric. An alternate procedure for these materials will be discussed later.

1. Air and/or Oven Drying:
   Many samples can be air-dried or oven dried (40-50°C) without fabric alteration or craking. Soils with high clay or organic matter contents are generally not well suited to air or oven drying. In soils with high levels of dissolved salts, crystals may form upon drying. Air-drying is relatively slow, and it may take considerable time for large samples to lose water. In any event, air-dried samples should be heated to 40- 50°C for 24-48 hours to remove hydroscopic water.

2. Acetone Replacement:
   Acetone replacement of water should be used in soils that have a high clay content, show evidence of shrink-swell potential, or where preservation of the voids, channels, and cracks is important. This is probably the best method to mimimize the formation of shrinkage cracks. Instead of allowing the water to evaporate form the sample, acetone is used to desplace the water by mass flow and diffusion. Water mixes well with acetone, and as the acetone becomes saturated with watr, it is periodically replaced until all the water is removed from the sample. With acetone replacement, the Kubiena tins or moisture cans are useful as they provide support for the sample while it is moist. The easiest procedure is to submerge the sample in pure actone. With Kubiena tins, both covers are removed, gauze fixed to the bottom, and the sample placed in stainless steel trays. Acetone is added until the sample is covered. A petri dish with anhydrous calcium chloride can be added as a dehydrant to speed up the drying process. The container should be sealed to prevent acetone evaporation, the procedure should be carried out in a fume hood. The acetone is replaced every 5 to 7 days with fresh acetone and repeated until there is no water in the acetone. This procedure may take several weeks or several months depending on the size and initial moisture content of the ample. The presence of water in the acetone can be checked by mixing 10 ml of the acetone solution with 40 ml of kerosene, or turpentine. Acetone is miscible with the kerosene, but water is not. If water is present, the solution will become turbid or the solutions will form a boundary layer.

3. Freeze Drying:
   Freeze drying is not commonly used, but may have some advantages in certain situations. In soils with abundant tree roots, it may not be possible to extract a sample because of all the roots. Freezing the sample permits one to cut out a block without disrupting the integrity of the sample (Blevins et al., 1970). The samples are placed in liqued nitrogen, or liquid nitrogen can be used to freeze the soil in situ. Using insulated boxes, the sample can be brought back to the lab without thawing. The ice that formed in the sample may freeze more slowly and form ice crystals. Where practical, other techniques should be used to dry the sample.

Except for very highly indurated samples, most soil material needs to be hardened by impregnation with a resin so the sample will retain its integrity whil being cut and polished. The ideal resin should be clear, isotropic, refractive index about 1.54, thermally stable, nonreactive with the sample constituents or cleaning solutions, no volume change during hardening, able to penetrate all pores (low viscosity), no health hazard, readily available, and inexpensive. We would be pleased to find such a product. Until the ideal product comes along, we will have to make do with the products that are available. there are two widely used products: polyester resin and epoxy resin. Each has advantages, limitations, and situations where it is most suited. Another product, Carbowax 8000, is used in special situations, but has some major limitations. these resins should be used with caution. Mixing should be done in a fume hood, and gloves and protective clothing and eyewear worn while preparing the resins and during the impregnation process.

1. Polyester Resins:
   Polyester resin has been used for a number of years by many scientists, and with success. The resin, by itself, is uaually too viscous to penetrate the smaller soil proes, and is generally diluted with actone or styrene. Although styrene is good choice, it does have health hazards and should be used with caution. Styrene also has a short shelf life. Acetone is generally recommended as a diluent because of its lower toxicity, and is ideal for samples that have undergone actone replacement. Dilutions of 2:1 polyester:acetone or 1:1 ratios are often used. Polyester resins also mix well with fluorescent dyes, such as Uvitex OB (CIBA-GEIGY). Uvitex OB is a UV fluorescent dye and is used in th estudy of pores. The fluorescent dye is not as useful in sandy soils because the abundance of packing voids cause the whole sample to fluoresce, making it difficult to differentiate the structural voids and cracks. This dye has bee used successfully in Vertisols to distinguish the pore distribution.

   The hardening of polyester resins is promoted by adding a small amount of a hardening agent, usually methyl ethyl ketone peroxide. The amount of hardener controls the hardening time. However, only a small amount of hardener is added so that the sample takes about 4-6 weeks to gel. This allows time for the rein to fully penetrate even the smalles pores. Final curing is accomplished by heating in a 40-50°C oven for 24-48 hours. Too much hardenere causes the sample to heat (exothermic reaction). An alternate procedure is to omit the hardener, or only use a very small amount, and promote hardening by exposing the sample to gamma irradiation. Commonly a Co radiation source is used. More recently we have used residual gamma radiation during shutdown of a research nuclear reator. We have used both radiation methods, and the samples are generally quite hard. If they are soft in spots, heating to 40-50°C for 24 hours will generally harden the sample.

   Although the impregnation and hardening procedures seem straight forward, problems may occur for apparently unknown reasons. Gremlins come in at night to sabotage the solution. Sometimes what works on one sample will not work on another. It is best to experiment on mixing ratios, hardeners, and dyes before doing a major project.

2. Epoxy Resins:
   Epoxy resins are generally more expensive than polyester resins, expecially the ones formulated for petrographic work. Thus, their use has been limited. However, they do fill a need and work well in some siturations, especially when you are in a hurry. They are generally more stable under an electron beam. Most epoxies do not mix with water or acetone. this make epoxy almost impossible to use for samples that had water displaced with acetone. Samples with acetone replacement are usually clayey soils. Epoxy impregnation seems to work better on coarser txuted soils. A sales representative of CIBA-GEIGY indicated that some of their Araldite resins will mix with samll amount of water and can be diluted with actone. You will need to check this on your own, as there is little information on this topic. Shell makes a clear resin that hardens at room temperture, has rather low initial viscosityk, and can be diluted with their own shell product. Other epoxies cannot be thinned withdiluents, and some requier heating to reduce viscosity and promote hardiening. Most epoxy resing have an exotermic reatic as they harden which causes the sample to heat. Generally this heating is not a problem unless you are impregnating a large volume.

   Some epoxies require heating to about 70-80°C to reduce viscosity and promote hardening. NOTE- heat the two parts separately and do not mix until ready to use. Gelling and hardening is an exothermic reaction, and heating only speeds up the reaction. Thus, we have a trade off; the higher the temperature the lower the viscosity, but the faster the gelling time. You need a solution that will penetrate the pores, but not harden before the sample is completely impregnated. Some trials with different temperatures will help you decide on the appropriate temperature. Because the viscosity is temperature dependent, the sample and the epoxy need to be at the same temperature. Because the viscosity is temperature dependent, the sample and the epoxy need to be at the same temperature. Heating a sample to 70-80°C may cause fabric alteration, cracking, mineral crystallization, or mineral transformation. Temperature sensitive samples should not be impregnated with epoxy requiring heat to lower the viscosity. One advantage of epoxy resins is that they gel quickly, usually within 34-48 hours. Heat activated epoxy samples placed back in the oven will harden within 24-48 hours. EPO-TECK makes a series of epoxies with low viscosity at room temperature, and cure overnight at room temperature. You may wish to investigat some of these other epoxies to find one that fits you needs and budget.

3. Carbowax 8000:
   Carbowax 8000 is polyethylene glycol with a molecular weight of 7000-9000. It is a white powder, or flakes, with a melting point of 60-63°C and solubility in water of 63 wt% at room temperature. It is relatively safe, and has been approved as a food additive. For perparing thin sections it has the disadvantage of being soft, low melting point, and crystallizes as birefringent radiating spherulites. Carbowax 8000 is useful in impregnating wet organic soils that cannot be dried with other techniques. In practice, the wet sample is immersed into the molten solution and retained at an elevated temperature of about 70°C for several days to a week. An alternate procedure is to prepare aqueous solutions of Carbowax 8000 ranging from 25%, 50%, 75%, to 100%. The sample is first placed is the 25% solution for several hours, then placed in successively more concentrated solutions. Murphy (1986) gives directions for this procedure.

   Because Carbowax 8000 is soft and water soluble, it creates problems with cutting and grinding. Water cannot be used as a lubricant, and sawing and grinding must be done carefully so as not to generate heat from friction. In wet organic soils, the matrix and the impregnant are about equal hardness, and can be cut with ease. In soils with abundant mineral matter, such as quartz, the soil matrix is harder than the Carbowax and grain plucking may occur. Samples impregnated with Carbowax 8000 need to be handled with care.

   IMPREGNATING the SAMPLE

   Selecting an impregnating solution and drying the sample is the first step. Having the solution enter the sample is another. there are various methods and equi9pment used, but most involve vacuum impreganation. This is true for polyester resins and epoxy, but with slight variations. The first step is to place the sample in a leak proof container. The container should be just slightly larger than the smaple, such that the sample is covered with 1-2cm of solution. For trimmed clods, glass tumblers or small drinking glasses work well. If you want to reuse the glass container, pick a design that is tapered (wider at the top) and without an upper lip. After drying, the sample will usually come out without breaking the glass. An easier means is to use disposable 250 or 400ml polyethylene beakers. these can also be reused, but not as often. If the container is much larger than the sample, you will be wasting solution. Samples collected in moisture cans will fit into pint paper containers. The container is first lined with heavy duty Al foil to prevent leakage. Kubiena tins may be placed in disposable loaf pans, or you may have to make your own container. Ther is no problem cutting the metal later.

   1. Polyester Resin:
      For polyester resin impregnation, place the samples in a vacuum desiccator and apply a vacuum. The vacuum is needed to evacuate air from the sample and replace it with a resin which will hold the sample together. the vacuum chamber should have a separate inlet to allow the addition of the impregnating solution while maintaining a vacuum. A vacuum can be drawn from a vacuum pump or water aspirator. If a vacuum pump is used, you will need to have a freeze and moisture trap to prevent vapors from contaminating the pump oil. After about 15 minutes under vacuum, slowly add the polyester resin down the side of the container while maintaining vacuum. this causes the solution to saturate the bottom of the sample first. Samples collected in moisture cans should have holes punched in the bottom to allow the entry of the resin. A strong boiling of the impregnating solution may occur. this is not the release of air from the sample, but acetone and other organic volatilization. If this occurs, turn off the cacuum source temporarily, or adjust the vacuum. It may take 10 to 30 minutes to add the resin solution to cover the sample. After the sample is covered, maintain the vacuum for about 30 minutes.

      Slowly release the vacuum and place the samples in a fume hood. They may be covered for a few days, but then uncovered to allow the acetone to evaporate. Samples will not harden properly unless the acetone has evaporated. Alternatively, place the sample in a pressure chamber after removing from the vacuum chamber. Pressures of about 100 psi for 24 hours should force the resin into any unfilled pores. Check the samples periodically to see if the resin solution has dropped. If it has, add additional resin to keep the sample covered with solution. Clean up all materials as soon as possible with acetone and dispose of cleaning materials properly. As a precaution, gloves, protective clothing, and protective eyewear should always be worn. A high vacuum in a glass container has the potential to cause damage if the glass breaks. We have never had an accident, but it pays to be cautious. If a hardener was added to the resin, gelling should be complete in about 6 weeks. Final curing is promoted by heating the sample at 4-50°C for several days. An alternate hardening procedure is to expose the impregnated sample to gamma irradiation either from a Co source or research nuclear reactor.

   2. Epoxy Resin:
      The procedure for epoxy resin is similar for polyester resin, but with some modifications. For resins that have to be at an elevated temperature to reduce viscosity, the solution should be added rather quickly. If the resin were added as described above, it would cool too fast. The two parts of the resin are removed form teh oven and mixed. Wear oven gloves to remove the resin from the oven. As soon as the resin is mixed, the samples are removed from teh oven and the resin poured over the samples so that they are covered. The samples are then placed in the vacuum chamber for about 30 minutes. If possible, you could heat the vacuum chamber to help maintian a low viscosity of the epoxy. After impregnation, the samples are topped off, and placed back in the oven for curing. they are usually hard within 24 hours. Epoxy resins that cure at room temperatures should be mixed just prior to impregnation as the mixing of the two components initiates the hardening process. The resins that cure at room temperature often have an exothermic reaction that promotes the polymerization process. Cooling the resin may extend its shelf life, but increase its viscosity.
   FINAL SAMPLE PREPARATION

   Assuming you have come this far successfully, you still have a ways to go. The cured sample is removed form the container, placed in a paper pint container, and the container is marked withthe appropriate sample notation (including orientation).

   1. Sawing and Trimming the Sample:
      Sawing is done with a rock-cutting saw using a diamond cut-off blade. There should be a shield on the blade for protection, and to prevent coolant from splashing onto the operator. Protective eyewear, earwear and clothing should be worn, as this can be a messy operation. If there is strong atomization of the oil during cutting, a mask should be worn or the saw placed in a fume hood. The coolant keeps the blade cool, which prevents warping, and flushes the cuttings away form the sample. In cutting the sample, use light pressure on the saw blade. Excess pressure will cause the blade to heat and warp, cutting grove into the sample. A light cutting or flushing oil is generally used. Some people have allergic reactions to petroleum products. If so, gloves can be worn, or use a coolant that is not petroleum based. Do NOT use kerosene as a cooant. It has a low flash point and th evapors can be ignited from sparks produced when cutting hard materials. Water should also be avoided because it may cause some soils to swee, especially clayey soils. The saw should have a vice to hold the smaple firmly as it is being cut, and to move the sample laterally in a fixed orientation.

      Commonly vertical and horizontal sections are prepared. If there is a special feature to be investigated, you may want to cut that feature befor making other sections. Clamp the sample in the vice so that the blade will cut about 1 cm into the bottom or side of the sample. Discard this cut section, as it may have been disturbed during sampling. Move the sample about 1-2 cm and make another cut. The inside of this slice will be the one mounted for a thin section. Move the sample 1-2 cm if additional sections are needed. If the smaple is moved without unclamping, both faces should be parallel. This aspect will be important later. Remove the sample and prepare a vertical section as described above. The cut slices can now be trimmed to fit the microscope slide. If possible, trim off any excess resin, as there is no need for it. You will want to polish and grind the impregnated soil slice, not resin. If you cut through a Kubiena tin or a moisture can, remove the metal as you trim the excess resin. The metal is soft and may cause problems when polishing the smaple. Mark the orientation on the back of the slice and place the cut pieces in the marked container. Proceed to cut additional samples until you have about 10-12 slices, or the number that will fit a bonding jig.

   2. Polishing the Sample:
      For a high quality thin section, the impregnated slice needs to be ground uniformly smooth and flat. The finished thin section will be about 30 µm thick, thus even small imperfections of a few µm will be noticed. Polishing is done with a lap wheel using lose powders, grinding paper, or by machine. If grinding powders or paper are used, begin with 120 or 240 grit and proceed to 400 and 600 grit. Each finer grit size removes scratches or marks introduced by the coarser grit. Check at each stage for smoothness and flatness. The finer grit sizes are for polishing only, as they will not remove much of the sample. Polishing should be done using a clean light cutting oil, not water. During the polishing, move the slice around to prevent cutting in one place. If you use one portion of a lap wheel, it may become dished out and no longer be flat. The slice should be carefully cleaned between each grinding operation to prevent coarser grit from contaminating a finer grit size. After the final polish, the slice should be carefully cleaned to remove all oil and grinding grit. Ethanol can be used to remove the oil. Be careful, as some resins may become soft with prolonged exposure to ethanol. Wipe the slice carefully with a tissue to help remove the oil and grit. Then, carefully wipe the slice with lens tissue or a non-lint tissue. Facial tissues are absorbent, but very linty, and any lint in the block will be readily observed in thin section. At this, and subsequent stages, cleanliness is very important.

   3. Bonding the Slice to the Glass Slide:
      The polished slice is bound to the microscope slide with epoxy resin formulated for that purpose. Do not use tube epoxy bought from the hardware store. Carefully mix the epoxy so as not to incorporate air bubbles. The microscope slide should be cleaned and wiped with lens tissue to remove any dust or oil. If the slide is scratched or appears to have imperfections, discard it. If you will be using a thin section machine to prepare the final section, you may want to pregrind the microscope slides to a common thickness. A box of microscope slides may be flat and smooth, but thickness is not uniform. One manufacturer lists the 50 x 75 mm slides as 1.21 ±0.005mm, another at 1.1 to 1.3mm. Unless the slides are ground to a common thickness, machine processing becomes more time consuming.

      A thin layer of clear epoxy is appliced to the slice, and the slide carefully placed on the slice in a hinge motion so that the epoxy is slowly pushed ahead of teh slide. After the slide is in place, carefully manipulate the slide to remove any air bubbles, and press out excess ipoxy. Place the mounted slide in a bonding jig. The purpose of the jig is to produce a uniformly thin layer of epoxy between the microscope slide and the slice. To have a uniform thickness of epoxy, the two surfaces of the slice should be parallel so there is uniform pressure over the area of the section. Maintain pressure on the slice until the epoxy hardens. If you do not have a bonding jig, use weights or a clamp. You can even make your own, or have one made at a local machine shop. Before the slice is placed on the bonding jig, place a piece of smooth Al foil on the base and coat the foil with a thin layer of oil (WD-40). Without the Al foil, you stand the chance of cementing the slide to the bonding jig.

   4. Finishing the Section:
      After the epoxy hardens (about 24 hours), remove the section from the bonding jig and carefully clean off excess epoxy, especially from the back of the microscope slide. If you have access to a thin section machine, place the section in the chuck and cut off as much of the impregnated slice as possible without disturbing the final 30 µm of the section. Normally, about 0.5mm of the section is retained on the slide. Carefully clean the section and transfer it to the final lapping portion of the machine. In small increments, begin removing the excess material. If the slides were preground to a uniform thickness, you may be able to polish to 30 µm with minimal difficulty. If not, the section needs to be examined frequently with a petrographic microscope to determine its thickness. The interference color of quartz, or some other mineral, will indicate the thickness of the section.

      If you do not have access to a thin section machine, you will need to use your cutoff saw to remove the excess material. Clamps are available which will hold a microscope slide in th esaw. You will probably not be able to saw off as much as the thin section machine, and will have to do more hand grinding. Grinding to the final thickness requires going through a seccession of finer grit powder or paper until you reach the final thickness. In either case, any final hand polishing can be done on glass plates using fine grinding powder.

      The finished section is carefully cleaned to remove oil and grinding powder. The slide is then marked with the profile, horizon, depth, lab number, an dorientation using India ink. Permanent cover slips can be attched using special epoxy or other mounting media. The purpose of the cover slip is to protect the surface of teh slide. If the section is to be stained for selective mineral identification, it should be cleaned of all oil and left uncovered until the staining is complete. A temporary cover slip can be attached using a viscous mounting medium, immersion oil, or th esection can be sprayed with a lacquer coating. In any case, for viewing the section, there needs to be a coating of oil or a cover slip to reduce light diffusion and produce a good quality image.

      CONCULSIONS

      this has been a brief attempt to acquaint you with the procedures, techniques, and mechancis of preparing a thin section. the thoughts and procedures are largely based on the experience of the author. If you want to pursue this subject further, you are encouraged to look at the references listed on a separate handout. Murphy (1986) does an excellent job of presenting material on thin seciton preparation. You are encouraged to read this book. Th efinal step is observing the section in the petrographic microscope, describing and interpreting the features observed. Texts by Brewer (1976), Bullock et al (1985) and Fitz Patrick (1993) are useful fo rdescribing your sections. (NOTE- Books by Brewer and Bullick et al are currently out of print). Other applications such as examining the section with a SEM or image analysis can also be used.

      Preparing thin secitons takes time, patience, and practice, especially for the beginner. You may want to check with your geology or engineering departments to see if they have any equipment or expertise in this area. Maybe they would be willing to share equipment and advice. GOOD LUCK.

      REFERENCES
      1. Blevins, R.L., N. Holowaychuk, and L.P. Wilding. 1970. Micromorphology of soil fabric at tree root-soil interface. Soil Sci. Soc. Am. Proc. 34:460-465.
         
      2. Brewer, R. 1976. Fabric and mineral analysis of soils. Robert E. Krieger Pub. Co., Huntington, New York.
         
      3. Bullock, P. et al. (eds). 1985. Handbook for thin section description. WAINE Research Publications. Albrighton, Wolverhampton, U.K.
         
      4. FitzPatrick, E.A. 1993. Soil microscopy and micromorphology. John Wiley & Sons, New York.
         
      5. Murphy, C.P. 1986. Thin section prepartion of soils and sediments. AB Academic Publishers. Berkhamsted, Herts, U.K.

      ---

      

      ---

      This page (http://www.ces.ncsu.edu/plymouth/programs/drees.html) created by
      Vera MacConnell, Research Technician, I on October 26, 1997.
      Last Updated on November 3, 1997.