Low Thermal Mass Gas Chromatography: Principles and Applications

Journal of Chromatographic Science, Vol. 44, May/June 2006

Low Thermal Mass Gas Chromatography: Principles and Applications Jim Luong1, Ronda Gras2, Robert Mustacich2, and Hernan Cortes3 1Dow

Chemical Canada, P.O. BAG 16, Highway 15, Fort Saskatchewan, Alberta, Canada, T8L 2P4; 2RVM Scientific, Inc., 5511 Ekwill St. # A, Santa Barbara, CA, 93111-2398; and 3Dow Chemical USA, Midland, MI, 48667

Abstract In gas chromatography (GC), temperature programming is often considered to be the second most important parameter to control, the first being column selectivity. A radically new GC technology to achieve ultrafast temperature programming with an unprecedented cool down time and low power consumption has recently become available. This technology is referred to as low thermal mass GC (LTMGC). Though the technology has its roots in resistive heating, which forms the basis of principle and design concept, the approach taken to achieve ultrafast heating and cool down time by LTMGC represents a significant break-through in GC. Despite some rectifiable shortcomings, LTMGC has proven to be an ideal methodology to deliver near/real time GC data, high precision, and high throughput applications. It is a new approach for modern high-speed GC. This paper documents the fundamental design principles behind LTMGC, performance data, and examples of applications investigated.

Introduction The process of increasing column temperature during a gas chromatographic (GC) analysis is referred to as temperature programming (TPGC). In GC, temperature programming is often considered to be the second most important parameter to control, the first being column selectivity (1–3). For a particular solute, TPGC leads to a decrease in retention volume and retention factor. The benefits of TPGC include better separation for solutes with a wide boiling range, improved detection limit, and improved peak symmetry, especially for solutes with high retention factors (1). In addition, TPGC is essential for the removal of unwanted heavier materials that might otherwise compromise the integrity of a chromatographic system. A new, recently introduced instrument incorporates technology to achieve ultrafast temperature programming with an unprecedented cool down time and a power consumption of approximately 1% of conventional GC. The technology is referred to as low thermal mass GC (LTMGC). In addition, this LTMGC technology is designed as an accessory for commercially available GCs. * Author to whom correspondence should be addressed: email [email protected].

This paper documents the fundamental principles behind LTMGC, performance data, and examples of applications investigated using the RVM LTMGC accessory (RVM, Santa Barbara, CA) for the Agilent HP-6890 series GCs (Wilmington, DE).

Experimental An RVM LTMGC accessory for Agilent HP-6890 GC series model LTM-A68 was used. The LTM-A68 was integrated with an Agilent HP-6890A GC equipped with a vent exhaust deflector, Transcendent Enterprise heated pressurized liquid injector (HPLIS)–pressurized liquid injector (PLIS), split/splitless injector, programmable temperature vaporizer injector, flame ionization detector (FID), and Valco pulsed discharge detector (VICI, Houston, TX). Column modules used for evaluation and applications developments included: (i) 18-m × 0.25-mm i.d., 0.25-µm VF-1MS column on a 3-inch standard tray; (ii) 5-m × 0.10-mm-i.d., 0.12-µm CP-Sil 8 CB column on a 3-inch standard tray; (iii) 7-m × 0.53-mm-i.d. CP-Lowox column on a 5-inch wide tray; (iv) 2-m × 0.10-mm-i.d., 0.12-µm CP-Sil 8 CB column on a 5-inch wide tray; and (v) 5-m × 0.10-mm-i.d., 0.12-µm DB-1 column with no transfer line on a 5-inch wide tray (Varian, Middleburg, the Netherlands). In all the GC applications described, hydrogen was used as the carrier gas, and a flame ionization detector operated with 30 mL/min hydrogen, 25 mL/min nitrogen, and 350 mL/min air was used as a detector. The GC conditions for the system suitability test shown in Figures 1–3 (Figure 3, see page 5A) are as follows: the LTMGC module used was a Varian VF-1MS column (18 m × 0.25-mm i.d., 0.25 µm). The LTMGC temperature profile was 40°C held for 1 min then increased 100°C/min to 150°C. The host oven temperature was 200°C. The format tray was a standard 3-inch tray. The average linear velocity was 50 cm/s. The injector was operated in split mode (50:1), and the injector temperature was 250°C. The transfer line was fused silica, uncoated, but deactivated. The interface was SGE zero dead volume connector (Victoria, Australia). The detector temperature was 300°C, and the injection was carried out manually at 1 µL.

Reproduction (photocopying) of editorial content of this journal is prohibited without publisher’s permission.

253


The GC conditions for the separation of 300 ppm (w/w) nC14–nC16 in hexanes (Figures 4 and 5) were as follow: the LTMGC module was a CP-Sil 8CB column (2 m × 0.10-mm i.d., 0.12 µm). The LTMGC temperature profile was 50°C for 0.5 min then increased 100°C/min to 300°C. The host oven temperature was 250°C. The format tray used was a standard 3-inch tray. The average linear velocity was 100 cm/s, and the injector was operated in split mode (50:1). The injector temperature was 250°C. The transfer line was fused silica, uncoated, but deactivated. Agilent Press-fit connectors were used as the interface. The detector temperature was 300°C, and the injection was carried out manually at 1 µL. The GC conditions for the analysis of volatile alcohols in hydrocarbons and stacked injection (Figures 6 and 7) were as follow: the LTMGC module used was a Varian Lowox column (7 m × 0.53-mm i.d.). The LTMGC temperature profile was 50°C held for 1 min then increased 300°C/min to 325°C and held for 1 min. The host oven temperature was 250°C. The format tray used was a 5-inch wide tray. The average linear velocity was 100 cm/s, and the injector was operated in the split mode (3:1). The injector temperature was 250°C. The transfer

Figure 1. Chromatogram of a Grob’s Test Mixture by LTMGC–FID. Note the excellent chromatogram obtained. For the figure: 1-octanol, 1; n-undecane, 2; 2,6-dimethylphenol, 3; octanoic acid methyl ester, 4; 2,6dimethylaniline, 5; naphthalene, 6; n-dodecane, 7; 1-decanone, 8; n-tridecane, 9; and decanoic acid methyl ester, 10. The concentration was 0.1% each in cyclohexane.

Figure 2. An overlay of chromatograms of chlorinated phenols with temperature programming rates from 200°C/min to 400°C/min. For the figure: phenol, 1; 2,4-dichlorophenol, 2; 4-chlorophenol, 3; 2,4,5-trichlorophenol, 4; and 2,4,5,6-tetrachlorophenol, 5. The concentration was approximately 300 ppm (v/v) each in hexane.

254

line was fused silica, uncoated, but deactivated. Restek Pressfit connectors were used as the interface. The detector temperature was 300°C. The injection was made with a PLIS valve. The GC conditions for the extractable chlorinated hydrocarbons analysis, (Figure 8) were as follow: the LTMGC module was a Varian VF-1MS column (18 m × 0.25-mm i.d., 0.25 µm). The LTMGC temperature profile was 40°C held for 1 min then increased 450°C/min to 250°C. The host oven temperature was 250°C, and the format tray used was a standard 3-inch tray. The average linear velocity was 200 cm/s. The injector was operated in the split mode (20:1). The injector temperature was 250°C. The transfer line was fused silica, uncoated, but deactivated. The interface used was an SGE zero dead volume connector. The detector temperature was 300°C, and the injection was performed manually.

Figure 4. An overlay of chromatograms nC14–nC16 with temperature programming rates of 100°C/min, 300°C/min, and 600°C/min. For the figure: nC14, 1; nC15, 2; and nC16, 3. The concentration was 300 ppm (w/w) each in hexane.

Figure 5. An overlay of chromatograms of nC14–nC16 with temperature programming rates of 1200°C/min, 1500°C/min, and 1800°C/min. For the figure: nC14, 1; nC15, 2; and nC16, 3. The concentration was 300 ppm (w/w) each in hexane.


The GC conditions for the analysis of organo-metallic tins (Figure 9) were as follow: LTMGC module was a Varian VF-1MS column (18 m × 0.25-mm i.d., 0.25 µm). The LTMGC temperature profile was 40°C held for 1 min then increased 450°C/min to 250°C. The host oven temperature was 280°C. The format tray used was a standard 3-inch tray. The flow velocity was 200 cm/s, and the average linear velocity was split (5:1). The injector temperature was 300°C. The transfer line was fused silica, uncoated, but deactivated. The interface used was an SGE zero dead volume connector, and an FID was used as the detector. The detector temperature was 300°C. The injection was performed manually. The GC conditions for the analysis of chlorinated phenols (Figure 10) were as follow: the LTMGC module was a Varian VF-1MS column (18 m × 0.25-mm i.d., 0.25 µm). The host Figure 6. A chromatogram of methanol, ethanol, and 2-propanol in hexane oven temperature was 280°C. The LTMGC temperature profile by LTMGC–FID. For the figure: methanol, 1; ethanol, 2; and 2-propanol, was 40°C for 1 min then increased 300°C/min to 250°C. For the 3. The concentration was 100 ppm (v/v) each in hexane. variable rate for the heating rate test, the temperature programming varied from 200°C/min to 400°C/min. The format tray used was a standard 3-inch tray. The average linear velocity was 60 cm/s. The injector was operated in split mode (20:1), and the injector temperature was 300°C. The transfer line was fused silica, uncoated, but deactivated. The interface was an SGE zero dead volume connector. The detector temperature was 300°C, and the injection was carried out manually. The GC conditions for the analysis of Norpar 12 (Figure 11) were as follow: the LTMGC module CP-Sil 8CB column (2 m × 0.1-mm i.d., 0.12 µm). The LTMGC temperature profile was 40°C held for 0.5 min then increased at a variable rate to Figure 7. Stacked injection of 10 ppm (v/v) each of methanol, ethanol, and 2-propanol in hexane by 300°C. The host oven temperature was HPLIS–LTMGC–FID. Top chromatogram: stacked of seven injections. Bottom chromatogram: stacked of 250°C. The variable rate was from five injections. 100°C/min to 1000°C/min. The format tray used was a 5-inch wide tray. The average linear flow velocity was 100 cm/s, and the injector was operated in the split mode (10:1). The injector temperature was 300°C. The transfer line was fused silica, uncoated, but deactivated. Press-fit connectors were used as the interface. An FID was used as the detector at a temperature of 300°C. The injection was carried out manually. GC conditions for simulated distillation analysis (Figure 12, see page 5A) were as follow: the LTMGC module was a Varian VF-1MS column (18 m × 0.25-mm i.d., 0.25 µm). The host oven temperature was 300°C. The LTMGC temperature profile was 40°C held for 0.5 min then increased 300°C/min to 300°C and held for 2 min. The format tray used was a standard 3-inch tray. The average linear flow velocity was 200 cm/s, and the injector was operated in the split mode (20:1). The injector temperature was 300°C. The transfer line was fused silica, Figure 8. A chromatogram of extractable chlorinated hydrocarbons by uncoated, but deactivated. An SGE zero dead volume conLTMGC–FID. For the figure: tetrachloroethane, 1; hexacloroethane, 2; nector was used as the injector. The detector temperature was hexaclorobutadiene, 3; pentachlorobenzene, 4; and hexachlorobenzene, 300°C, and the injection PLIS. 5. The concentration was 200 ppm (w/v) each in hexane. Standards used for testing were obtained from Aldrich

255


Chemicals (Milwaukee, WI) and VWR Scientifics Products (Edmonton, Alberta, Canada), and other chemicals were obtained from various production plants at Dow Chemical Canada (Western Canada Operations, Fort Saskatchewan, Alberta, Canada).

100°C/min can be attained. Cool down time, however, is very long especially when the initial temperature approaches ambient temperature. Another shortcoming with this design is that the power consumption reaches the kilowatt level for a single temperature cycle. Miniature heating elements with GC columns have been proven to be the most successful approach for achieving rapid temperature programming at rates of up to 1200°C/min (3–8). Results and Discussion Some typical approaches include resistively heated metal coated columns, resistively heated metal walled columns, Principle and design of LTMGC smaller heater plates, internal heater wire, and a resistively Temperature programming is often conducted by electronic heated element with a column in a sheath. temperature control of an oven that houses the GC column. In theory, resistively heated temperature programming is The relatively small thermal mass of the capillary GC column quite attractive. The fundamental principle behind it is rather along with the winding of the column on a wire frame support straightforward, based primarily on Ohm’s law, and the hardprovides extensive surface contact of the column with the ware involved is inherently space conscious. Despite this heated air for fast and reproducible temperature equilibrium. discovery in the early 1950s (9), the deployment and commerUsing this type of arrangement and depending on the size of cialization of said technique has been protracted. Figure 13 the oven, reproducible temperature programming of up to shows a chronology of at-column resistive wire heating. Electrical shorting is the major concern of this principle, exacerbated by the inclusion of additional metallic components such as a heating wire, resistive temperature detector (RTD), or metal columns in such close proximity to each other. The lack of an appropriate insulation material that can withstand elevated temperatures and temperature cycles has further aggravated the development (9). In 2001, Mustacich et al. was granted two patents for the invention of LTMGC (10,11). This led to the recent commercialization of a radically new disruptive technology to achieve ultrafast temperature programming and unprecedented cool down time with a power consumption of only approximately 1% of conventional GC (12–16). In the context of GC operation, the terminology low thermal mass (LTM) is used to qualitatively describe GC column assemblies, including both heating and temperature sensing means that have a very small total mass compared with GC designs that are Figure 9. A chromatogram of organo-metallic tins by LTMGC–FID. For the heated by convection means. For temperature programming of figure: tetratethyl tin, 1; tetrabutyl tin, 2; and tetrapentyl tin, 3. The concentration was 100 ppm (w/w) each in hexane. the GC column, this results in a very small mass being heated, hence the LTM description for GC designs of this type. For the LTM approach to be useful, a certain level of thermal efficiency is implicit, otherwise the GC column assembly will not easily heat even though it has a low total mass. The LTMGC technology achieves a high heating efficiency through its reduction of the surface area of the assembly. At the heart of LTMGC is the column module assembly. A typical module assembly consists of a capillary column—columns of any dimension or length can be used; a 2-m platinum resistive temperature detector (RTD) with typical diameter of 0.02 to 0.03 inches; a nickel alloy heating wire with a typical diameter of 0.08 inch; a metal tray to support the module, transfer lines, and for heat dissipation; and a microprocessorFigure 10. A chromatogram of chlorinated phenols by LTMGC–FID. For the figure: phenol, 1; controlled electric fan to facilitate rapid heat 2,4-dichlorophenol, 2; 4-chlorophenol, 3; 2,4,5-trichlorophenol, 4; and 2,4,5,6-tetrachlorophenol, removal during the cooling down cycle of a run 5. The concentration was 100 ppm (w/w) each in hexane. that is mounted at the bottom of the metal tray.

256


To employ ultralow thermal mass technology, a highly precise entire assembly is then twisted into a torus and covered with a RTD is a necessity in the design. The oven compartment in a thin aluminum foil as external skin. Figure 14 shows a diagram conventional GC is relatively massive (i.e., it has a large specific of the assembly and a cross section of the module. heat). If the specific heat of the system is large enough relative The low thermal mass design, coupled with the high surface to the rates of loss mechanisms, namely convection and conarea, offers a theoretical rapid temperature programming rate duction, and is relatively constant, it is certainly possible to of up to 30°C/s, or 1800°C/min, in addition to an unprecedirectly control the delivery of heat and employ a sensorless dented cooling down time. Power consumption is approxisystem. One must also be confident in such a system that the mately 1% that of conventional GC, primarily because of the distribution of heat is not spatially localized because the column effective heating of interstitial air with the random positioning and heating element are not the same component and heat is of the heating wire in the torus and the low thermal mass of either convected or conducted to the column according to the the assembly (10). geometry and design of the apparatus. By this geometry, fast Heat capacity is the quantity of heat required to increase the temperature equilibrium and temperature repeatability are temperature of a system or substance by one degree of temperafacilitated, even though there may be a difference in the actual ture. It is usually expressed in calories or Joules/°C relative to a temperature of the column itself. Even if the size of the oven is significantly reduced, such as in the case of many small GCs, the thermal mass is still much larger than the convective and conductive losses. If the convective and conductive losses can be maintained relatively constant through the design of the apparatus, then it may still be possible to use a sensorless design. In the LTM design, the oven is completely eliminated, and the mass of the system has collapsed down to the capillary tubing, fine wires, micron-sized insulation fibers, and thin aluminum wrapping. Unique to the LTMGC design, the specific heat is a minute fraction of the convective and the conductive heat loss. These losses are relatively insignificant when compared with a convection oven. The specific heat may now only be a few percent of the total power. The overall power requirements have dropped approximately 100 times when compared with an oven for temperature proFigure 11. An overlay of Norpar 12 by LTMGC–FID with temperature programming rates of gramming. This high efficiency results in a 100°C/min, 500°C/min, and 1000°C/min. The concentration was 100 ppm (w/w) Norpar 12 in cross-over such that natural convection is hexane. responsible for most of the power consumption. This process is highly variable and is a function of many variables such as local flows, convection currents, temperature gradients around the periphery of the instrumentation, surface area, assembly packing density, specific length of column in the assembly, type of column, specific wire lengths, and types of wire alloys, as well as others. This variability means it is not possible to obtain repeatable heating results with a specific amount of power delivered to the heating element. For this reason, an ultralow thermal mass, high precision RTD and an advanced control algorithm are combined with the column for accurate and highly precise temperature measurement in LTMGC (14–16). The column module utilizes an advanced composite material, commercialized by 3M (St. Paul, MN), to rove the RTD/assembly and then the heating wire with a specially designed weaving device. This insulating material advantageously meets the contrary disparative objectives of preventing shorting between Figure 13. The history of at-column resistive wire heating. Diagram courtesy components while maximizing the heat conducted between of Dr. Leslie Ettre. the heating element and these components. After roving, the

257


Figure 14. A diagram of the LTMGC assembly and a cross section of the module. Picture courtesy of Dr. Robert Mustacich, inventor of LTMGC technology.

unit mass or system mass. For the LTMGC, the system mass is small and, therefore, the system heat capacity, based on the heat capacities of the components, is relatively small. This resulted in a very small quantity of heat required to raise the temperature of the system when compared with standard GC systems. The impressive low consumption of power by LTMGC is clearly illustrated in Figures 15 and 16, respectively. In Figure 15, a conventional GC was subjected to a temperature programming run of 40°C held for 0.5 min then increased 30°C/min to 180°C and held for 3 min. The average power required for this programming run was approximately 1090 W. In Figure 16, an LTMGC module with a capillary column (1 m × 0.25-mm i.d., 0.25 µm) was subjected to a temperature programming run of 40°C held for 0.35 min then increased 60°C/min to 180°C and held for 0.35 min. The average power required for LTMGC was less than 5 W. In this specific case, a reduction of power consumption of approximately 200 times was realized. The significance of an ultralow power GC is yet to be appreciated. LTMGC not only has the potential to change how GC is being conducted in different analytical theatres such as laboratories, in-situ, near line, or online, but it also has the potential of becoming a very important technology enabler to other analytical techniques being practiced. On the basis of principle and design concept, the technology has its root in resistive heating; however, the approach taken to achieve ultrafast heating and cool-down time represents a significant breakthrough in GC. Performance

Figure 15. Power profile of a conventional GC. Temperature profile: 40°C held for 0.5 min then increased 30°C/min to 180°C and held for 3 min. The average peak power consumption was 1.09 kW.

Figure 16. Power profile of LTMGC. Temperature profile: 40°C held for 0.35 min then increased 60°C/min to 180°C and held for 0.35 min. The average peak power consumption was 5 W.

258

Installation The LTM-A68 accessory, built on the principle of LTMGC, can accommodate up to four different column modules in standard format, two column modules in standard format and one column module in wide format, or two column modules in wide format. These modules can be operated independently from one another. Up to eight methods, each method with a maximum of 10 temperature programming ramps, can be stored by the on-board microprocessor. Installation involves simply replacing the door of the host oven with the LTM-A68 controller/door and the modules are then connected to the appropriate injector/detector with fusedsilica transfer lines. The host oven and LTM operation are independent of each other and can be operated as such. No modification of the electronic system is required for the host GC to integrate with the LTM-A68, apart from the attachment of a remote controlled cable. A system set up is relatively straightforward and takes less than 1 h to complete. In Figure 17 (see page 6A), the modules attached to the oven door of the conventional gas chromatograph contain two different columns: (i) a Varian VF-1MS column (18 m × 0.25mm i.d., 0.25 µm) in a 3-inch standard tray or (ii) a Varian CPLowox (7 m × 0.53-mm i.d.) in a 5-inch wide tray. System suitability The LTM-A68 can be interfaced to any injector or detector via deactivated fused silica column transfer lines. To demonstrate the performance of the unit, a Varian VF-1MS (18 m × 0.25-


mm i.d., 0.25 µm) in a standard tray was used for evaluation. Interfacing was carried out using two approximately 100-cm deactivated 0.25-mm i.d. fused-silica retention gaps. A standard mixture consisting of 0.1% each of 1-octanol, n-undecane, 2,6-dimethylphenol, octanoic acid methyl ester, 2,6-dimethylaniline, naphthalene, n-dodecane, 1-decanone, n-tridecane, and decanoic acid methyl ester in cyclohexane was analyzed. The GC conditions used were listed in the Experimental section, and Figure 1 shows a chromatogram obtained. Excellent peak symmetry for all probe compounds was observed, highlighting the high degree of inertness of the chromatographic system under the conditions used. Heating LTMGC fast temperature capability was tested using the module described in System suitability section. This module was not ideal for ultra fast temperature programming because of the fact that it has a rather long column, but it was useful to demonstrate the practical aspect of LTMGC being used in conventional settings and conditions. Figure 2 shows an overlay of chromatograms of chlorinated phenols with temperature programming rates varied from 200°C/min to 400°C/min. Compression of peak capacity can be seen even at 400°C/min. Figure 3 shows an overlay of chromatograms with temperature programming rates of 300°C/min, 400°C/min, and 500°C/min; however, little gain was observed. Clearly, for this module, the maximum temperature-programming rate is around 400°C/min. To demonstrate the ultrafast temperature programming capability of LTMGC, a CP-Sil 8CB module (2 m × 0.1-mm i.d., 0.12 µm) on a wide tray was used. Figures 4 and 5 show overlays of chromatograms of a test solute comprising 300 ppm (w/w) each of nC14, nC15, and nC16 hydrocarbons in hexane with temperature programming rates ranging from 100°C/min to 1800°C/min. The GC conditions used were listed in the Experimental section. Even at 1800°C/min, peak capacity compression was observed. This fast programming rate makes this module ideal for use with hyphenated techniques such as comprehensive liquid chromatography–GC or GC–GC. Key learnings from these experiments suggest that although LTMGC can be programmed at a rate of up to 1800°C/min, its maximum rate was a function of thermal mass of the column assembly. That is, for a 2-m × 0.1-mm-i.d. column, 1800°C/min can be reached without difficulty; whereas for an 18-m × 0.25mm-i.d. column, a limit was reached somewhere between 400°C/min and 500°C/min. Cooling Like heating, in LTMGC, cooling rate is also inversely proportional to the column module thermal mass (i.e., a faster cooling rate is attained when the column module mass is smaller). In addition, the cooling rate is also a function of the amount of surface area and the rate of heat being removed by mechanical fans. Figure 18 (see page 6A) shows a comparison of cooling rates of four modules to that of an Agilent HP6890A GC equipped with a vent deflector. As can be seen from the conventional Agilent HP-6890 data, it takes slightly more than 11 min to cool down from 300°C to

30°C at 22°C ambient temperature. For LTMGC, with the higher thermal mass module such as that of the 18-m column, it takes 4 min. With the shorter columns of 2 to 7 m in length, an average cooling time of only 1.3 min was obtained using LTMGC. To cool from 300°C to 30°C, depending on the thermal mass of the modules, LTMGC was 3 to 10 times faster when compared with conventional GCs. For a CP-Sil 8CB column (2 m × 0.1-mm i.d., 0.12 µm) in a wide tray format, the time required to cool from 250°C to 50°C was an unprecedented 25 s without the use of a cryogen, making the module ideal for use in the second dimension of comprehensive multidimensional chromatography or where high throughput or near real time data (or both) are required. Applications

The following are some industrial applications investigated using the LTM-A68 unit. Alcohols in hydrocarbons Volatile alcohols, such as methanol, are specification items in hydrocarbon final products. The presence of methanol in hydrocarbons such as ethylene had a negative consequence on catalyst performance. The special features of LTMGC in rapid heating and cooling, combined with a highly selective column, makes it suitable for improving throughput and sensitivity. Figure 6 shows a chromatogram of 100 ppm (v/v) of methanol, ethanol, and 2-propanol in hexane with a total analysis time of approximately 3.5 min. Hexane was used mainly for ease of handling. The GC conditions used were listed in the Experimental section. Table I shows the repeatability of retention time and area counts obtained. Very respectable chromatographic performance was achieved. Less than 0.1% relative standard deviation (RSD) at 95% confidence level for retention time and less than 3% at 95% confidence level for area counts was observed despite the fact that the system was programmed at a very fast rate of 400°C/min. Stacked injection, a novel concept of employing a column stationary phase for both sample enriching and a separating medium to improve solute detectability, has been described earlier (17,18). The special features of LTMGC were advantageous for this type of analysis by using a low initial temperature to remove the matrix and the very fast temperature programming capabilities to elute the trapped solutes and improve sensitivity and throughput. Figure 7 shows an overlay of stacked injections of five and seven injections of 10 ppm (v/v) of methanol, ethanol, and 2-propanol in hexane. The chromatogram shows five solvent peaks for five injections and seven solvent peaks for seven injections, correspondingly. Methanol, ethanol, and 2-propanol of these injections were focused into three discrete peaks, respectively. When compared with conventional GC, a throughput improvement of at least five times was realized when using LTMGC (18). Extractable chlorinated hydrocarbons Extractable chlorinated hydrocarbons such as 1,1,2-tricholoethane, 1,1,2,2-tetrachloroethane, hexachlorobutadiene, pentachlorobenzene, and hexachlorobenzene are routinely

259


analyzed in environmental applications. Total analysis time required for this application, including cool down time, was approximately 35 min (19). Figure 8 shows a chromatogram of the same sample using LTMGC. Cool down time included, a total analysis time of 3 min was attained, representing an analytical throughput improvement of 10 times. The GC conditions used were listed in the Experimental section. Organo-metallic tins Organo-metallic tins are commonly used as anti-coking agents in various industrial applications. Close monitoring of the solutes was required to determine dosing efficiency. With conventional GC, the total analysis time was approximately 30 min. Figure 9 shows a chromatogram of organo-metallic tins in hexane using LTMGC. A total analysis time of less than 3 min was attained, representing an analytical throughput improvement of 10 times when compared with conventional GC. The GC conditions used were listed in the Experimental section.

ASTM-D-2887 with nC44 eluting at approximately 2 min and a total analysis time of approximately 3 min. This represents an overall throughput improvement of 10 times over conventional GC. The GC conditions used were listed in the Experimental section. Limitations/constraints Some limitations and rectifiable constraints include: the host oven has to be kept constantly at an elevated temperature. This might have a negative impact on the capillary tubing used

Retention time (min)

Chlorinated phenols Figure 19. A chromatogram of Norpar 12 by conventional GC. The conditions were: GC, Agilent HP-6890N; split/splitless injector in split mode; Chlorinated phenols such as chloro-, di-, tri-, tetra-, and pen15:1 ratio; helium carrier at 21 cm/s; column, CP-Sil5CB MS (30 m × 0.32tachlorophenol are commonly analyzed in environmental applimm i.d., 1 µm). The temperature program was 120°C held for 0 min then cations by GC. With conventional GC, total analysis time was increased 20°C/min to 250°C and held for 4 min. The concentration was approximately 40 min (20). 100 ppm (w/w) Norpar 12 in hexane. Figure 10 shows a chromatogram of chlorinated phenols in methanol using LTMGC. A total analysis time of less than 3.5 min was attained, repTable I. Reproducibility of Retention Times and Area Counts for 100 ppm (v/v) resenting an analytical throughput Each of Methanol, Ethanol, and Propanol in Hexane by LTMGC–FID improvement of approximately 10 times. The GC conditions used were listed in the Retention time (min) Integration area counts Experimental section. Run

Norpar 12 Exxon Norpar 12 is a paraffin based hydrocarbon fluid used for ethylene recovery. Figure 19 shows a chromatogram of Norpar 12 by conventional GC. A total analysis time of approximately 12 min was required. Figure 11 shows an overlay of chromatograms of Norpar 12 by LTMGC using different temperature programming rates from 100°C/min to 1200°C/min with a total analysis time of approximately 2 min. Clearly, with LTMGC, a six times faster analysis can be conducted. The Experimental section listed the GC conditions used. Simulated distillation Simulated distillation analysis is often carried out in petroleum industries as per ASTM-D-2887 for product characterization. A typical analysis, using conventional GC, takes approximately 30 min for the elution of hydrocarbons of up to nC44. Figure 12 (see page 5A) shows an overlay of chromatograms of two standards for

260

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Average SD* %RSD 95% CV†

Methanol

Ethanol

Propanol

Methanol

Ethanol

Propanol

1.891 1.893 1.891 1.891 1.892 1.890 1.892 1.892 1.891 1.892 1.891 1.892 1.891 1.891 1.891 1.891 1.891 1.891 1.892 1.891 1.891 0.0007 0.0355 0.08

1.953 1.952 1.953 1.952 1.952 1.952 1.951 1.951 1.951 1.952 1.952 1.953 1.952 1.953 1.952 1.952 1.952 1.953 1.953 1.952 1.952 0.0007 0.0344 0.08

2.001 2.002 2.001 2.001 2.000 2.001 2.002 2.001 2.001 2.002 2.001 2.002 2.001 2.002 2.001 2.002 2.001 2.002 2.001 2.002 2.001 0.0006 0.0293 0.06

50.3 50.4 49.1 49.7 49.8 48.7 48.9 50.4 50.1 50.4 50.9 50.1 50.9 50.9 50.1 50.5 50.7 50.9 50.6 50.1 50.175 0.656 1.3075 2.9

90.1 90.2 89.1 89.2 89.1 90.3 90.1 90.3 90.1 90.2 92.1 90.0 91.0 90.7 91.3 92.1 91.9 90.8 91.2 91.0 90.54 0.9005 0.9946 2.2

83.9 83.4 83.4 83.2 83.1 83.2 83.1 83.2 83.1 83.5 84.5 84.1 83.9 83.2 83.2 83.2 84.9 82.7 83.9 84.5 83.56 0.5807 0.695 1.5

* SD = standard deviation. † 95% CV = standard deviation at the 95% confidence interval.


for interfacing the host GC with the LTMGC. For example, if the carrier gas used contains moisture or air, this might lead to undesired reactions such as hydrolysis or stationary phase oxidation resulting in excessive bleed. Because the host oven has to be kept constantly at an elevated temperature, if LTMGC is used with a cool-on column injection system or with a splitless injector, an uncoated yet deactivated transfer line (retention gap) should be employed to connect the injector with the column module so that proper chromatographic focusing effect can take place at the inlet of the column module. With the host oven being at a higher temperature than the LTMGC module, especially when the system is in standby mode, chromatographic impurities such as septum bleed, stationary phase decomposition, or impurities in carrier gas can accumulate in the LTMGC module. This requires thermal conditioning of the LTMGC prior to analytical work being conducted. The interfacing of the LTMGC to the host oven must be conducted in such a fashion that chromatographic fidelity is not destroyed. Mixing, cold spot, or band broadening can occur if connection between the analytical columns and transfer lines are not done properly. Although fan and assembly trays are user-exchangeable, the column module assemblies are not. Columns, regardless of vendor sources, can be coiled only by the manufacturer. This requires careful planning and an on-hand inventory of appropriate column modules in applications that are deemed critical.

Conclusion Though the technology has its root in resistive heating, the approach taken to achieve ultrafast heating and cool down time by LTMGC represents a significant breakthrough in GC. Innovations in the domain of oven design and thermal management resulted in a fast heating rate of up to 1800°C/min and unprecedented cool down time in seconds. LTMGC is ideal for use in applications that require near or real-time analysis and is a mission critical component for high-speed GC. The ultralow power consumption of LTMGC not only has the potential to change the way GC is currently being practiced, it may be a potent technology enabler to other analytical techniques in analytical chemistry. In the applications evaluated, when compared with conventional GC, LTMGC shows markedly improved chromatographic performance and sample throughput. With the advent of LTMGC, new techniques such as stacked injection (18), pyrolysis LTMGC, LTMGC×LTMGC can be developed to further enhance the applicability of GC in analytical chemistry.

Acknowledgments Special thanks to Dr. James Griffith, Dr. Terry McCabe, Rony Van Meulebroeck, Patric Eckerle, Myron Hawryluk, Lyndon

Sieben, Vicki Carter, and Dr. Don Patrick for their contributions in the implementation of LTMGC technology. Credit must be given to Dr. Mary Fairhurst and the Leveraged Technology Separations Leadership Team for their support of this project. The authors would also like to express their appreciation to the editors and reviewers for their assistance and advice in preparing the manuscript. This project was partially funded by Analytical Science’s Corporate Innovation Funds.

References 1. M. Van Deursen. “Novel Concepts for Fast Capillary Gas Chromatography”, Ph.D. Thesis, ISBN 90-386-2873-0, CIP-Data Library Technische Universiteit Eindhoven, Eindhoven, 2002. 2. H. McNair and J. Miller. Basic Gas Chromatography. John Wiley and Sons, Inc., New York, NY, 1997. 3. C. Poole. The Essence of Chromatography. Elsevier Science B.V., Amsterdam, the Netherlands, 2003. 4. V. Jain and J. Phillips. Fast temperature programming on fused silica open tubular capillary columns by direct resistive heating. J. Chromatogr. Sci. 33: 55–59 (1995). 5. D. Ehrmann, H.P. Dharmasena, K. Carney, and E.B. Overton. Novel column heater for fast capillary gas chromatography. J. Chromatogr. Sci. 34: 533–39 (1996). 6. E. Overton and K. Carney. New horizons in gas chromatography: field applications of microminiaturized gas chromatographic techniques. Trends Anal. Chem. 13(7): 252–57 (1994). 7. W. Maswadeh. “New generation of hand-held compact disposable gas chromatography devices”. Workshop on Field-Portable Chromatography and Spectrometry, Snowbird, UT, June 3–5, 1996. 8. H. Dubsky and M. Fatscher. Step programmed temperature GC. J. Chromatogr. 47: 297–306 (1970). 9. L.S. Ettre and R. Mustacich. Personal communications. 2001. 10. R. Mustacich and J. Everson. U.S. Patent 6,217,829, April 17, 2001. 11. R. Mustacich and J. Richards. U.S. Patent 6,209,386, April 3, 2001. 12. R. Mustacich. U.S. Patent 5,782,964, July 21, 1998. 13. R. Mustacich and J. Richards. U.S. Patent 6,490,852, December 10, 2002. 14. R. Mustacich, J. Richards, and J. Everson. U.S. Patent 6,530,260, March 11, 2003. 15. R. Mustacich and J. Everson. U.S. Patent 6,682,699, January 27, 2004. 16. R. Mustacich, J. Everson, and J. Richards. Fast GC: thinking outside the box. Am. Lab. March: 38–41 (2003). 17. J. Luong, C. Mork, L. Sieben, and B. Winniford. “A novel approach for trace analysis of oxygenated compounds in light hydrocarbons”. Plenary Lecture #12, Proceedings from the 23rd International Symposium on Capillary Chromatography, Riva Del Garda, Italy, June, 2000. 18. J. Luong, R. Gras, H. Cortes, and R. Mustacich. Stacked injection with low thermal mass gas chromatography for PPB level detection of oxygenated compounds in hydrocarbons. J. Chromatogr. Sci. 44: 219–26 (2006). 19. U.S. Environmental Protection Agency. Analysis of Extracted Chlorinated Hydrocarbons Method, 3rd ed. SW-846 publication. No. 955-001-00000-1. U.S. Government Printing Office, Washington, D.C., 1980. 20. U.S. Enivornmental Protection Agency. Analysis of Chlorinated Phenols in Water Method, 3rd ed. SW-846 publication. No. 955-00100000-1, U.S. Government Printing Office, Washington, D.C., 1980. Manuscript received September 7, 2005; revision received February 25, 2006.

261


Low Thermal Mass Gas Chromatography: Principles and Applications

Figure 3. An overlay of chromatograms of chlorinated phenols with temperature programming rates of 300°C/min, 400°C/min, and 500°C/min. For the figure, 300°C/min (green trace), 400°C/min (red trace), and 500°C/min (blue trace). From left to right: phenol, 2,4-dichlorophenol, 4-chlorophenol, 2,4,5-trichlorophenol, and 2,4,5,6-tetrachlorophenol. The concentration was approximately 300 ppm (v/v) each in hexane.

(see pp. 253–61)

Figure 12. An overlay of nC8–nC44 hydrocarbons (simulated distillation standards) by LTMGC–FID.

5A


Figure 17. A Picture of LTM-A68 interfaced with an Agilent HP-6890A GC. In this picture, an LTM-A68, shown connected to an Agilent HP6890A GC, is equipped with two column modules: (top) a Varian CPLowox (7 m × 0.53-mm i.d.) in a wide tray and (bottom) a Varian VF-1MS column (18 m × 0.25-mm i.d., 0.25 µm) in standard tray.

6A

Figure 18. A comparison of cool down rates between an Agilent HP-6890A GC and various LTMGC modules. A comparison of cooling rates between conventional GC to LTMGC modules. RVM-1-ST: Varian VF-1 column (18 m × .25-mm i.d., 0.25 µm). RVM-2-ST: Varian CP-Sil8 CB (5 m × 0.1-mm i.d., 0.12 µm). RVM-3-WT: Varian Lowox (7 m × 0.53-mm i.d.). RVM-4-WT: Varian CP-Sil8 CB (2 m × 0.1-mm i.d., 0.12 µm). For the figure, Blue: oven temperature program: 30°C held for 0.5 min then increased 30°C/min to 250°C and held for 1 min. Red: oven temperature program: 40°C held for 0.5 min then increased 30°C/min to 250°C and held for 1 min.