Crystal Growth, Selection and Mounting
Table of Contents
Dr. Paul Boyle's Crystal
Growing Recipes--an excellent resource. Other ideas for growing crystals
are at the Bijvoet
Center's web site at the University of Utrecht. A page by
at the University of Nottingham
describes a variety of different techniques for growing crystals.
Information about growing crystals of
compounds is available at the Cambridge University web site. A commercial supplier
with a variety of hints and tools to grow protein crystals is
Producing good quality crystals of a suitable size is the first and most
important step in determining any crystal structure. Crystallization is the
process of arranging atoms or molecules that are in a fluid or solution state
into an ordered solid state. This process occurs in two steps--nucleation
and growth. Nucleation may occur at a seed crystal, but in the absence of seed
crystals usually occurs at some particle of dust or at some imperfection in the
surrounding vessel. Thus if crystals do not appear to come out of supposedly
supersaturated solutions, try to
seed the crystallization by introducing
either micro crystals from a previous attempt or by rubbing the glassware to
induce imperfections in the glassware. Adding foreign particles
to the solution to induce crystallization is not recommended.
The shapes of crystals depend on both the internal symmetry of the material and on the relative growth rates of the faces. In general, the faces of the crystal that grow most rapidly are those to which the crystallizing particles are bound more securely. These rapidly growing faces are usually the smaller, less well developed faces. The larger faces are usually associated with directions in the crystal where there are only weak intermolecular interactions.
All crystallization methods change the physical state of a material by transforming the system from some non-equilibrium state toward an equilibrium state. Crystallization methods may be separated into two broad categories based upon how the system performs this transformation. Concentration gradient methods typically involve concentrating the sample by either removal of solvent or transport of the material to another solvent system in which the material is less soluble. Thermal gradient methods rely upon the fact that crystals form when a material is cooled.
The choice of crystallization method for a particular sample depends greatly upon the physical and chemical properties of the sample. Properly choosing the best solvents, crystallizing agents, and temperatures is essential to producing top quality crystals.
There are a few general points that apply to all crystallization methods.
It is important that the sample be as pure as possible. When crystallization attempts consistently yield oils, the sample is probably not pure. The solvents or cocrystallizing materials should also be as pure as possible.
It is important for most solution methods that the glassware be thoroughly clean. Note that new glassware often has dusty contaminants, so it must be thoroughly cleaned especially before the first use.
If a sample only yields small crystals, the method should be altered so as to slow down the growth step. Slowing the crystal growth sometimes requires changing the method used to grow the crystals.
Avoid vibrations near your growing crystals as this moves the system to an equilibrium state faster than desired. See previous point.
When growing crystals by some concentration gradient method, use the smallest amounts of solvent needed to dissolve the sample.
Finally be patient! Some methods work in a few hours, and other methods require weeks or even months for success.
Concentration Gradient Methods
Evaporation is by far one of the easiest methods for crystallizing small molecule compounds. The choice of solvent is very important because it can greatly influence the mechanism of crystal growth and because the solvent may be incorporated into the crystalline lattice. It is customary to screen a large number of solvents or solvent mixtures to find the best conditions for crystal growth. The rate of crystal growth can be slowed either by reducing the rate of evaporation of the solvent or by cooling the solution. Formation of only a few rosette-shaped clumps is an indication of an insufficient number of nucleation sites. The number of nucleation sites may be increased either by seeding the solution or by scratching the surfaces of the vessel exposed to the solution. Finally, do not, under any circumstances, use solvents that contain a large number of different compounds such as petroleum ether or gasoline. See information about the choice of solvent below.
Liquid and Vapor Diffusion
Liquid and vapor diffusion methods are often tried when evaporation methods do not immediately succeed. Both methods require finding two solvents or solvent mixtures in which the compound is soluble in one solvent but insoluble in the other solvent. The two solvent systems should be immiscible or nearly immiscible for liquid diffusion and should be miscible for vapor diffusion. Crystal growth may be slowed somewhat by cooling the apparatus.
Liquid diffusion usually requires that the less dense solvent system be carefully layered on top of the more dense system in a narrow tube. The sample can be dissolved in either solvent system. Crystals grow at the interface between the solutions. When compounds precipitate immediately upon being formed, it is possible to slow down the reaction and thus grow larger crystals by putting the reactants in different liquid layers that are separated by a third solvent layer that is not miscible with either of the layers or with the sample. Note that the top layer should be added very slowly to assure a minimum of mixing of the layers. These tubes can be checked for crystal growth by placing a light behind the tube and looking for the shiny facets of the crystals. Crystals of air- and moisture-sensitive samples can grown in Schlenk-ware tubes with a septum on top.
Vapor diffusion is carried out by dissolving a small amount of the sample in a small vial or tube, then placing this open small vial or tube inside a larger vial that contains a small portion of a solvent in which the sample is insoluble. The outer vial is then sealed. During crystallization, vapor from the solvent of the outer vial diffuses into the solution in the inner vial causing the material to precipitate. The vertical surfaces of the inner vial should not touch the outer vial to keep the outer solution from rising by capillary action and filling the inner vial.
Protein crystals are often grown using a variation of the vapor diffusion method. A drop of the protein in mother liquor is either placed on a cover slip that is sealed over a well with a precipitating solution or the drop is added to a separate well or ledge that has a vapor path to the precipitating solution.
Some compounds, that precipitate as very small crystals immediately upon synthesis, are extremely insoluble. Suitable crystals of these compounds can often be prepared by greatly decreasing the rate at which the reactants combine. This is done by making the reactants diffuse through a gel barrier. To do this, fill the bottom part of a U tube with a gel, then introduce the reactants in the two separate ends of the tube. Such methods usually take weeks to months to produce crystals, depending on the rate of diffusion of reactants through the gel.
Thermal Gradient Methods
Thermal gradient methods often produce very high quality crystals. Such methods include slow cooling of sealed, saturated solutions, refluxing of saturated solutions, sublimation, and zonal heating. Zonal heating is used primarily for crystallizing solid solutions or mixtures. Small crystals may sometimes be grown larger by zonally refluxing a saturated solution. Sublimation may be carried out in a variety of tubes or vessels. Sealed vessels offer an advantage for sublimation in that the chamber may be evacuated or a partial pressure of some inert gas may be introduced before sealing the sample in the apparatus. Sublimation methods consistently produce very high quality crystals. Larger crystals may be grown either by decreasing the thermal gradient or by cyclic heating and cooling of the sample.
Other Details about Crystallization
Cocrystals and Clathrates
The crystal structure of some compounds can only be determined by coordinating the compound of interest with another material or by incorporating the compounds into a lattice of another material. Crystals that contain two or more different compounds are called cocrystals. Some crystal mixtures are simply formed by the incorporation of one or more solvent molecules into the lattice of the compound of interest. Other cocrystal mixtures are formed when the compound of interest is bonded to a large molecule such as triphenylphosphine oxide usually through a hydrogen bond. A final group of cocrystals can be thought of as being formed by incorporating the compound of interest or guest molecule into the small vacant regions in the lattice around large, rigid host molecules. This lattice of host/guest molecules is called a clathrate. Structures of porphyrin-based clathrates are very common.
A new class of clathrate crystals has recently been described in Nature, 2013, 495, 461-466, doi:10.1038/nature11990. To prepare these materials, metal complex crystals with large voids (that contain an unbound solvent) are soaked with micro- or nano-gram quantities of a compound of interest. Hopefully the compound of interest enters the voids displacing enough of the solvent to be indetified in the resulting crystal structure.
Choice of Solvents and Counter Ions
There are a number of solvents and counter ions that are commonly found to be disordered in crystal structures and thus should be avoided when growing crystals, if possible. The solvents giving the most trouble are petroleum ether, mixed hydrocarbons like hexanes or kerosene, and halogenated hydrocarbons such as methylene chloride and chloroform. Often these solvents occupy sites in a crystal structure that are larger than the solvent molecule, and thus appear to assume a variety of orientations in the single unit cell under study. When a group of atoms assumes a variety of positions in a lattice, the group is described as lacking order or simply disodered. The halogenated solvents are particularly troublesome when they are disordered because the disorder usually includes atoms that are heavier than the bulk material. Better choices of solvents are benzene, xylene, primary and secondary alcohols, and tetrahydrofuran. The mixed solvents such as pet ether or hexanes should never be used; good substitutes are always available. If good quality crystals can only be grown using a halogenated solvent, then by all means use that solvent. Getting good quality crystals is the most important step in the whole crystal structure process.
The counter ions most likely to cause difficulties because of their propensity to disorder are Bu4N+, BF4-, and PF6-. Some alternative counter ions that are usually ordered are triflate, BPh4-, (Ph4P)2N+, and Ph4As+.
To evaluate the quality and appropriate size of crystalline samples, the samples should be examined under low power (10X to 40X) magnification. Good crystals usually have smooth flat faces, sharp edges, no inclusions, no striations, and no obvious dislocations. A crystal with inclusions is shown below. Careful notes should be made if the bulk sample is not visibly homogeneous. The selected crystal should show no obvious external twinning (e.g. reentrant faces or different parts of the crystal extinguishing at different rotation angles under a polarizing microscope). The color, dimensions, habit, and point group symmetry of the crystal selected for examination should be noted.
The crystal chosen for analysis needs to be large enough to produce an adequate diffraction pattern and, at the same time, relatively small to minimize absorption problems. The calculation of structure factor amplitudes assumes that the crystal is being completely bathed in a uniform beam of X-rays. Since the uniform region of the X-ray beam is about 0.5 mm in diameter, this is taken as the maximum recommended dimension of the crystal in the past. For most samples, a minimum dimension of 0.1 mm is needed to produce adequate X-ray scattering. Compounds with few atoms or very heavy atoms can have all three dimensions toward the small end of this 0.1 to 0.5 mm range. Crystals of compounds with many light atoms should have all three dimensions toward the large end (0.4-0.5 mm) of this range.
Note that the upper limit on the size of a crystal can be relaxed. Crystals much larger than the X-ray beam can produce good quality results. Carl Henrik Görbitz published an informative article documenting the effects of collecting data on crystals larger than the X-ray beam size.(Acta Cryst., 1999, B55, 1090-1098.)
If the crystals are strongly absorbing (contain many heavy atoms), it is worthwhile to reshape the crystal to make it as nearly spherical as possible. Cutting, grinding or dipping the crystal in solvent are the best methods for reshaping the crystal.
Crystal mountings must be rigid enough to hold the sample in a fixed orientation and must minimize the amount of extraneous material that is in the incident and diffracted X-ray beam paths. The sample support is usually made from an amorphous material such as glass or plastic that is held in a metal pin and clamped onto a goniometer head. Solid glass fibers may be used; however, fibers pulled from glass tubing are actually small capillary tubes and are more rigid than solid glass fibers. These narrow tubes also place less non-crystalline material in the X-ray beam path than solid fibers.
Air stable crystals are usually glued (using epoxy, Elmers/water, Duco/amyl acetate, etc.) to the end of a glass fiber. The sample should be mounted with its smallest surface attached to the end of the glass fiber to minimize absorption effects and to minimize background scattering from the sample mount.
Mildly air unstable compounds can be coated with epoxy or an inert viscous material such as Paratone N oil, available from Hampton Research. These mountings are usually carried out in an inert atmosphere such as a dish filled with argon gas. The crystal is further kept from reacting during data collection by cooling the sample in a chilled, inert (nitrogen) gas stream.
Very reactive compounds must be mounted in a glove bag or glove box. Crystals of these compounds may be mounted using an inert coating on the crystal as described above or may be mounted in glass capillaries. If capillaries are chosen as the sample support, the crystals may be wedged in place or may be held in place by a small amount of (stopcock) grease. Capillary tubes containing unstable compounds must be sealed by melting the ends of the glass tube.
Capillaries do introduce two kinds of problems. The curvature of the capillary distorts the image of the crystal when centering the sample on the diffractometer. Also, the glass itself significantly increases both the background scattering and the absorption of the incident beam of X rays. It is crucial that the capillaries be made out of thin glass similar to that found in commercially-available capillaries for crystallography. Thick glass capillaries absorb most, if not all, of the diffracted X rays.