In the most basic level, a capillary column is composed of two parts - the tubing and stationary phase (Figure 3). Fused silica and stainless steel are the primary tubing materials. There are numerous stationary phases. Most are high molecular weight, thermally stable polymers that are liquids or gums. The most common stationary phases of this type are the polysiloxanes (sometimes incorrectly called silicones) and polyethylene glycols. The next most common type of stationary phases are small, porous particles composed of polymers or zeolites (e.g., alumina, molecular sieves).
Figure 3. GC Capillary Column
Fused silica is a synthetic quartz of high purity. A protective coating is applied to the outer surface with polyimide being the most common coating material. The polyimide coating is responsible for the brownish color of fused silica capillary columns. The color of the polyimide coating often varies between columns. Column color has no effect on the chromatographic performance of the column. Polyimide coated tubing often darkens after prolonged exposure to higher temperatures. The upper temperature limit of standard polyimide coated fused silica tubing is 360?C. High temperature polyimide coated tubing has an upper limit of 400?C.
The inner surface of fused silica tubing is chemically treated to minimize interactions of the sample with the tubing. The reagents and process used depend on the type of stationary phase being coated onto the tubing. A silylation process is used for most columns. Silanol groups (Si-OH) on the tubing surface are reacted with a silane type of reagent. Typically, a methyl or phenyl-methyl silyl surface is created for most columns.
Stainless steel capillary columns are used for applications requiring very high column temperatures. Stainless steel tubing is more robust than fused silica tubing, thus it is also used in situations where the possibility of tubing breakage needs to be virtually eliminated. Stainless steel interacts with many compounds, thus it is treated to minimize these undesirable interactions. The inner surface is either chemically treated or it is lined with a thin layer of fused silica. When properly made, the inertness of stainless steel capillary columns rivals those made with fused silica tubing.
Polysiloxanes are the most common stationary phases. They are available in the greatest variety and are the most stable, robust and versatile. Standard polysiloxanes are characterized by the repeating siloxane backbone (Figure 4). Each silicon atom contains two functional groups. The type and amount of the groups distinguish each stationary phase and its properties. The four most common groups are listed in Figure 4.
Figure 4. Polysiloxanes
The most basic polysiloxane is the 100% methyl substituted. When other groups are present, the amount is indicated as the percent of the total number of groups. For example, a 5% diphenyl-95% dimethyl polysiloxane contains 5% phenyl groups and 95% methyl groups. The "di-" prefix indicates that each silicon atom contains two of that particular group. Sometimes this prefix is omitted even though two identical groups are present. If the methyl percentage is not stated, it is understood to be present in the amount necessary to make 100% (e.g., 50% phenyl-methyl polysiloxane contains 50% methyl substitution). Cyanopropylphenyl percent values can be misleading. A 14% cyanopropylphenyl-dimethyl polysiloxane contains 7% cyanopropyl and 7% phenyl (along with 86% methyl). The cyanopropyl and phenyl groups are on the same silicon atom, thus their amounts are summed.
For select stationary phases, a low bleed or "ms" version is available. These stationary phases incorporate phenyl or phenyl type groups into the backbone of the siloxane polymer (Figure 5). These types of stationary phases are commonly called arylenes. The phenyl group strengthens and stiffens the polymer backbone which inhibits stationary phase degradation at higher temperatures. This results in lower column bleed and, in most cases, higher temperature limits. The arylene stationary phase substitution can be adjusted to maintain the same separation characteristics as the original, non-arylene version (e.g., DB-5 , DB-5ms , HP-5 , or HP-5ms ). The separations for the two versions are the same or extremely similar. It is rare, but there may be a slight separation difference between the regular and low bleed versions of a stationary phase. There are some unique low bleed stationary phases with no "regular" equivalent.
Figure 5. Low Bleed Phases (Arylene)
Polyethylene glycols (PEG) are widely used as stationary phases (Figure 6). Stationary phases with "wax" or "FFAP" in their name are some type of polyethylene glycol. Polyethylene glycols stationary phases are not substituted, thus the polymer is 100% of the stated material. They are less stable, less robust and have lower temperature limits than most polysiloxanes. With typical use, they exhibit shorter lifetimes and are more susceptible to damage upon over heating or exposure to oxygen. The unique separation properties of polyethylene glycol makes these liabilities tolerable. Polyethylene glycol stationary phases must be liquids under GC temperature conditions.
Figure 6. Polyethylene Glycol
There are two types of polyethylene glycols in common use as GC stationary phases. One has a higher upper temperature limit ( DB-WAXetr and HP-Innowax ), but exhibits slightly higher activity (i.e., peak tailing for some compounds). The other has a lower upper temperature limit and a lower low temperature limit but exhibits better reproducibility and inertness ( DB-WAX and HP-WAX ). The separation characteristics of the two stationary phases are slightly different. Another variation of polyethylene glycol phases is pH modifi- cations. FFAP columns are terephthalic acid modified polyethylene glycols ( DB-FFAP and HP-FFAP ). These columns are used for the analysis of acidic compounds. Base modified polyethylene glycol stationary phases are also available for the analysis of basic compounds ( CAM and HP-Basic WAX ). Strong acids and bases often exhibit peak tailing for standard columns. pH modified stationary phases may decrease the amount of tailing for strong acids or bases.
Gas-solid stationary phases are comprised of a thin layer (usually <10 ?m) of small particles adhered to the surface of the tubing. These are porous layer open tubular ( PLOT and HP-PLOT ) columns. The sample compounds undergo a gas-solid adsorption/desorption process with the stationary phase. The particles are porous, thus size exclusion and shape selectivity processes also occur. Various derivatives of styrene, aluminum oxides and molecular sieves are the most common PLOT column stationary phases.
PLOT columns are very retentive. They are used to obtain separations that are impossible with conventional stationary phases. Also, many separations requiring subambient temperatures with polysiloxanes or polyethylene glycols can be easily accomplished above ambient temperatures with PLOT columns. Hydrocarbon and sulfur gases, noble and permanent gases, and low boiling point solvents are some of the more common compounds separated with PLOT columns.
Some PLOT columns may occasionally lose particles of the stationary phase. For this reason, using PLOT columns that may lose particles with detectors negatively affected by particulate matter is not recommended. Mass spectrometers are particularly susceptible to this problem due to the presence of a strong vacuum at the exit of the column.
Bonded and Cross-linked Stationary Phases
Cross-linked stationary phases have the individual polymer chains linked via covalent bonds. Bonded stationary phases are covalently bonded to the surface of the tubing. Both techniques impart enhanced thermal and solvent stability to the stationary phase. Also, columns with bonded and cross-linked stationary phases can be solvent rinsed to remove contaminants. Most polysiloxanes and polyethylene glycol stationary phases are bonded and cross-linked. A few stationary phases are available in an nonbonded version; some stationary phases are not available in bonded and cross-linked versions. Use a bonded and cross-linked stationary phase if one is available. All J&W columns with the prefix "DB" are bonded and cross-linked stationary phases.
Column bleed is the background generated by all columns. It is the continuous elution of the compounds produced from normal degradation of the stationary phase. Column bleed increases at higher temperatures. The appearance of column bleed is best illustrated with a bleed profile or trace, generated by a blank run using a temperature programmed run that reaches the column's upper temperature limit and holds at that temperature for 10-15 minutes. A typical bleed profile is shown in Figure 7.
Figure 7. Bleed Profile
There are several important characteristics to a bleed profile. The baseline is relatively flat during the lower temperature region of the blank run. A sharp rise in the baseline begins at 30-40?C below the upper limit of the column and continues until the upper temperature limit is reached. Upon holding at the upper temperature limit, a fairly flat baseline is obtained. Several minutes may elapse before the baseline becomes completely flat. Major deviations from this profile are not due to column bleed. Column bleed is a continuous process - it is not an occasional or momentary event that stops and starts. To obtain a peak a band of volatile material has to enter and travel through the column. Any peaks in a blank run are not from column bleed, but most likely originate from contaminants in the GC system. Even if a mass spectral library search indicates silicon based compounds, the peaks are not from the column stationary phase; they may originate from the septum.
On average, polar stationary phases have higher column bleed, and significant bleed occurs at lower temperatures. Column bleed appears to be higher when using a detector that is particularly sensitive to any of the atoms or functional groups in the stationary phase. While the column bleed is not higher, the detector's enhanced sensitivity to the degradation products results in a very large baseline rise. This is usually most prominent for cyanopropyl substituted polysiloxanes used with an NPD or polyethylene glycol columns used with an ECD.
Excessive column bleed appears as a larger rise in the baseline at the higher temperature regions. There is no absolute measurement to indicate when column bleed is excessive. Column bleed is best measured as the difference or change in the background signal at two temperatures. Usually the column's upper temperature limit and a lower value around 100?C are used (Figure 7). The absolute background signal is a composite of the background generated by the entire GC system. It is not possible to determine the contribution of column bleed to this total background signal. By measuring the relative amount of column bleed, the other contributors to the background signal are subtracted out. Most columns are tested using FIDs. The output signal for a FID is in picoamps (pA). Bleed levels are usually reported as the difference (DpA) in the FID signal at two temperatures. Because these values are influenced by the detector response they have validity in column comparisons only if both measurements use the same detector under the same conditions, or both detectors are standardized and the bleed reported as pg C per gram of stationary phase under standardized flow conditions.
Column bleed increases as a column is used. Exposing the column to oxygen (air) and/or consistently using the column at or near its upper temperature limit for prolonged periods accelerates the onset of higher column bleed. A sudden or rapid increase in column bleed is usually an indicator of column damage or a problem in the GC system. Prolonged heating of a column above its upper temperature limit, constant exposure of the column to oxygen (usually via a leak), or repeated injection of damaging compounds are the most common causes or problems.
Column Temperature Limits
Columns have lower and upper temperature limits. If a column is used below its lower temperature limit, rounded and wide peaks are obtained (i.e., loss of efficiency). No column damage has occurred; however, the column does not function properly. Using the column at or above its lower limit maintains good peak shapes.
Upper temperature limits are often stated as two numbers. The lower one is the isothermal temperature limit. The column can be used indefinitely at this temperature and reasonable column bleed and lifetime are realized. The upper number is the temperature program limit. A column can be maintained at this temperature for 10-15 minutes without severely shortening column lifetime or experiencing excessively high column bleed. Exposing the column to higher temperatures or for longer time periods results in higher column bleed and shorter column lifetimes. Exceeding the upper temperature limits may damage the stationary phase and the inertness of the fused silica tubing.
Column capacity is the maximum amount of a solute that can be introduced into a column before significant peak distortion occurs. Overloaded peaks are asymmetric with a leading edge. Overloaded peaks are often described as "shark fin" shaped (Figure 8). Tailing peaks are obtained if a PLOT column is overloaded. No damage occurs if a column is overloaded.
Column capacity is related to stationary phase polarity, film thickness, column diameter and solute retention. Higher column capacities are obtained for solutes that are similar in polarity to the stationary phase. For example, a polar column has a higher capacity for a polar solute than a non-polar solute. Thicker film and wider diameter columns have higher capacities. Column capacity decreases as solute retention increases, thus two similar polarity solutes may overload the column at different amounts with lower capacity for the later eluting solute. Capacity ranges for the most common column diameters and film thicknesses can be found here .
Figure 8.Overloaded Peak
Figure 9. Boiling Point Elution Order for Homologous Series