96 EQUIPMENT of tubing wall in Fig. 3.6b). With properly designed systems, sub-milliliter volumes of tubing can be used for each solvent, with degassing effectiveness equivalent to extended vacuum degassing under static conditions for most HPLC operations [5]. Although not quite as effective as helium sparging, for most users the con- venience of in-line vacuum degassing and the cost of helium have made this the preferred degassing technique in place of helium sparging. For applications where dissolved-oxygen concentrations are critical (e.g., some fluorescence and electro- chemical detection methods) or where extremely low and constant concentrations of dissolved gases must be maintained (e.g., maximum sensitivity refractive-index detection), continuous helium sparging followed by in-line vacuum degassing is the best choice [5]. 3.4 TUBING AND FITTINGS Tubing and the fittings used to connect tubing to various HPLC system components are required for transporting the mobile phase and sample through the chromato- graph. If reasonable care is taken in the selection and use of tubing and fittings, problems will seldom be encountered. However, improper selection and use can generate unwanted extra-column volume (Section 3.9), which can compromise the separation—especially for small peak volumes. Additional technical information on tubing and fittings can be found on the Internet [6–7] and in manufacturer’s literature, such as [8]. Note that the following discussion refers to the dimensions and thread sizes of tubing and fittings in English units (usually fractional or decimal inches), because this is the way they are commonly supplied in the United States. These products are available in metric sizes in many other markets. As a note of caution, if both English and metric versions of similar products are used, be sure to label them clearly—even if they may seem to fit together in some cases, since damage to the part or a leak may result. 3.4.1 Tubing 3.4.1.1 Low-Pressure Tubing For pressures less than ≈100 psi, polymeric tubing generally is suitable. The two primary applications are transport of the mobile phase from the reservoir to the pump, and waste from the detector to the waste container. On the inlet side of the pump, tubing generally is 1/8-in o.d. and 1/16-in or smaller i.d. The inertness of fluorocarbon tubing (e.g., Teflon ® ) makes it the first choice for transport of solvents to the pump. Other polymers (e.g., polypropylene and polyethylene) may be suitable as well, but these products are best purchased from an HPLC supplies vendor to ensure that they are of sufficient purity for the application. Teflon tubing is somewhat permeable to gas, so air can diffuse through the inlet tubing into the mobile phase. Generally, air is not a problem, but in some applications (e.g., reductive electrochemical applications) dissolved oxygen in the mobile phase can cause problems. PEEK (polyetheretherketone) tubing is another suitable inlet-line tubing; it is not gas-permeable but it is opaque, so it is not possible to see bubbles inside the tubing. PEEK tubing also has some limitations in terms of chemical compatibility (Section 3.4.1.2). 3.4 TUBING AND FITTINGS 97 On the waste side of the detector, Teflon, polypropylene, polyethylene, PEEK, or other relatively inert tubing can be used. For ease of connection with the detector outlet, 1/16-in o.d. tubing generally is used as a waste line. The use of tubing with an internal diameter of 0.010-in i.d. or less will create sufficient back-pressure to keep bubbles in solution until they exit the detector cell—this will help reduce bubble problems in the detector. However, with this technique, one should be careful not to overpressure the detector cell, especially if the flow rate is increased dramatically. It usually is more prudent to use larger i.d. tubing (e.g., ≥0.20-in) and a back-pressure restrictor (available from many HPLC fittings suppliers) at the outlet end of the tubing to maintain 50 to 75 psi back-pressure at any flow rate. The length of tubing is not critical in low-pressure applications, but it is a good idea to keep these lengths to a convenient minimum. Excessive tubing lengths can result in longer washout times and delayed equilibration. Most low-pressure tubing is cut with a razor blade or, for PEEK, a cutter supplied by the tubing vendor. A flat cut, perpendicular to the tube axis is desired. 3.4.1.2 High-Pressure Tubing Whereas low-pressure tubing is used primarily to transport mobile phase to the pump or waste from the detector, high-pressure tubing is required elsewhere in the system. Conventional HPLC systems are designed for use with pressures up to 6000 psi between the pump and detector, so the tubing must be able to withstand such pressures. Also tubing used to transport the sample from the autosampler to the column and from the column to the detector must be sufficiently inert that sample adsorption or degradation does not take place, and the tubing length and diameter should be selected so that it does not contribute significantly to peak broadening (Section 3.9). Both stainless steel and PEEK tubing will satisfy these requirements for most applications. Type 316 stainless-steel tubing is most commonly used for HPLC applications. It is inert to nearly all solvents and has a very high burst strength. It is less convenient to use than PEEK tubing because of its stiffness. PEEK tubing ≤0.030-in i.d. will work up to pressures of ≈7000 psi without rupturing; for higher pressures, stainless-steel tubing is required. PEEK tubing is compatible with most HPLC mobile phases, but it will swell and become brittle when exposed to THF, chlorinated solvents, or DMSO (see [6 or 7] for a full listing of solvent compatibility). It is best to avoid PEEK when these solvents are present. The convenience and flexibility of PEEK tubing make it a good choice for use when connections are changed regularly, such as between the autosampler and column, and column and detector. Whenever the mobile phase contains components corrosive to stainless steel, the use of PEEK tubing will help minimize problems. When connections are made once and then forgotten, such as between the pump and autosampler, stainless-steel tubing is a more trouble-free choice. Tubing with an outside diameter of 1/16-in is used in most HPLC equipment; for some applications, 1/32-in o.d. tubing is preferred, but it is more fragile. Typical internal diameters for 1/16-in o.d. tubing are listed in Table 3.1. Stainless-steel tubing is readily available in sizes from 0.005-in to 0.046-in i.d., whereas PEEK covers the 0.0025-in to 0.040-in range. The 0.010-in and 0.020-in i.d. sizes are used to transport solvent from the pump to the autosampler, where extra volume 98 EQUIPMENT Table 3.1 Internal Diameters for Common High-Pressure Tubing (1/16-in o.d) Internal Diameter (in) Internal Diameter (mm) a Volume (μL/cm) 0.0025 b 0.06 0.03 0.005 0.13 0.13 0.007 0.18 0.25 0.010 0.25 0.51 0.020 0.50 2.03 0.030 0.75 4.56 0.040 1.00 8.11 0.046 c 1.20 10.72 a Nominal value (calculated from inch dimensions). b Available only in PEEK. c Available only in stainless steel. is not a concern, and these diameters are unlikely to become blocked. In the tubing where sample is present—between the autosampler and the column, and between the column and detector—smaller i.d. tubing (e.g., ≤0.007-in i.d.) is needed to avoid excessive peak broadening. Larger diameter tubing (≥0.030-in i.d.) is used primarily for construction of injector loops, because of the relatively large volume per unit length (Table 3.1). Care should be taken to select the tubing length and diameter so as to minimize peak-broadening contributions. Table 3.2 shows the impact of various combinations of tubing length and diameter on peak broadening for several representative column configurations. It can be seen that although 0.007-in i.d. tubing is satisfactory for conventional columns of ≥100 × 4.6mmwith≥3-μm particles, smaller columns require smaller diameter tubing. Applications that use sub-2-μm particles and/or long tubing runs (e.g., LC-MS) will require the use of 0.0025-in i.d. tubing. The smaller the tubing, the more prone it is to blockage from particulates that originate from the mobile phase, pump seal, valve-rotor wear, or sample. Consequently special care must be taken to avoid blockage when using tubing of ≤0.005-in i.d Remember that the tubing lengths shown in Table 3.2 are the total of the autosampler-to-column plus column-to-detector connections. Sometimes it is advantageous to use a short piece of 0.007-in i.d. tubing between the autosampler and column, so as to minimize blockage problems, and to use ≤0.005-in tubing between the column and detector, to minimize peak broadening. See Section 3.9 for an additional discussion of extra-column peak broadening. Because it is difficult to duplicate the quality of factory-cut tubing, it is best to buy precut lengths of stainless-steel tubing. Precut tubing has the added advantage of having been thoroughly cleaned and passivated, so it can be used without further treatment. Bulk stainless-steel (type 316 is recommended) can be purchased for cutting to lengths that cannot be purchased precut. Stainless-steel tubing can be cut easily with a tubing cutter purchased from the tubing supplier. The tubing should be flushed with several milliliters of solvent prior to use (connect the up-stream end of 3.4 TUBING AND FITTINGS 99 Table 3.2 Guide to Tubing Length Column Characteristics Maximum Length (cm) for 5% Increase in Bandwidth a L (mm) d c (mm) d p (μm) N (h ≈ 3) 0.0025-in 0.005-in 0.007-in 150 4.6 5.0 10,000 1450 90 25 150 2.1 5.0 10,000 300 20 * 100 4.6 3.0 11,100 580 35 10 100 2.1 3.0 11,100 120 * * 100 2.1 1.8 18,500 70 * * 100 1.0 1.8 18,500 15 * * 50 4.6 3.0 5,500 290 20 * 50 2.1 3.0 5,500 60 * * 50 1.0 3.0 5,500 15 * * 50 4.6 1.8 9,200 170 10 * 50 2.1 1.8 9,200 35 * * 50 1.0 1.8 9,200 * * * ∗≤10 cm. a Conditions for linear velocity = 2.5 mm/sec within the column (e.g., 2 mL/min for a 4.6-mm i.d. column), k = 1. the tubing to the HPLC system and direct the outlet to waste), so as to remove any residual oils or particulate matter. PEEK tubing can be cut easily in the laboratory, so it usually is purchased in bulk. For best results, a PEEK tubing cutter is used to score a line around the tubing, then the tubing is flexed to snap it; this gives a higher quality tube end than cutting all the way through the tubing. 3.4.2 Fittings 3.4.2.1 Low-Pressure Fittings These fittings are used to connect tubing when the pressure will not exceed ≈100 psi. The two most common fitting designs are shown in Figure 3.7. The flared-tubing fitting (e.g., Cheminert ® ,Fig.3.7a) requires a special tool to flare the tube end. A washer is used between the nut and the flared end to help secure the tubing in the fitting port. These fittings require some skill to flare but are inexpensive and reliable, so they are popular with instrument manufacturers to reduce manufacturing costs. An alternate design uses a ferrule to secure the tube end (e.g., Fingertight ® , Fig. 3.7b). The ferrule is reversed from the normal orientation in high-pressure fittings (as in Fig. 3.8a) so that the flat end contacts the bottom of the fitting port. The nut contains an internal taper that matches the ferrule so that the ferrule is swaged onto the tubing when the nut is tightened. This type of fitting is easy to assemble (the knurled nut is tightened with finger pressure), which has made it a popular alternative to the flared-tubing fitting. The industry standard is to use 1 4 -in nuts with 28 threads per inch (1/4-28); this way low-pressure fittings from different manufacturers are interchangeable. 100 EQUIPMENT knurled, finger-tightened nut (b) flat-bottom port inverted ferrule (a) tubing nut washer flange (flared end) (c) union tube end inverted ferrule internal taper nut threads in here Figure 3.7 Fittings for low-pressure applications. (a) Cheminert ® fitting, using a washer and flange for the seal; (b) Fingertight ® fitting, using an inverted ferrule for the seal. (c) Low-pressure connection of two tube ends with inverted ferrule of (b) in a union (nuts not shown for clarity). (a) Courtesy of VICI Valco Instruments Co. Inc.; (b) courtesy of Upchurch Scientific, Inc., a unit of IDEX Corporation in the IDEX Health & Science Group. Tubing connections for low-pressure fittings are made as shown in Figure 3.7c. Here two tube ends with inverted-ferrule fittings are connected in a union; the two ferrules butt against each other to make the connection (compare this to the high-pressure union of Fig. 3.8b,c). In other applications (e.g., a solvent- proportioning manifold, Section 3.5.2.2) a flat-bottomed fitting port is formed in the mating piece, as shown in the partially assembled fitting of Figure 3.7b. Finger pressure is all that is needed to tighten low-pressure fittings, so it is recommended that all such fittings be finger tightened, even if the nut is designed for use with a wrench. When a wrench or pair of pliers is used to tighten the fitting, it is easy to overtighten the fitting, and distort it or damage the threads. For applications where the fitting might vibrate loose (e.g., on the solvent-proportioning manifold in low-pressure mixing systems), some manufacturers offer a lock nut to provide extra security for low-pressure fittings. Remember, in a low-pressure application sometimes a loose fitting will allow air to leak into the liquid stream without liquid leaking out, so a loose fitting does not necessarily create a puddle of mobile phase. 3.4 TUBING AND FITTINGS 101 (a)(b) (c) (d ) nut ferrule tube end knurled,fin g er-ti g htened nut taper por t Figure 3.8 Compression fittings for high-pressure tubing connections. (a) Conventional stainless-steel nut, ferrule, and tube end; (b) union body; (c) properly assembled union. (d) Finger-tightened PEEK nut, ferrule, and fitting port. Courtesy of Upchurch Scientific, Inc., a unit of IDEX Corporation in the IDEX Health & Science Group. 3.4.2.2 High-Pressure Fittings All high-pressure tubing connections are made with fittings that use a ferrule to secure the tube end (Fig. 3.8a) in the fitting port (Fig. 3.8b). The fitting body, whether it is a union, check valve, or other fitting, contains a threaded portion for the nut, a taper for the ferrule, and a cylindrical port where the tube end contacts the fitting. To assemble the fitting, the nut and ferrule are slipped over the tube end, the tube end is inserted into the fitting until it bottoms out in the port, and then the nut is tightened to secure the fitting. When the nut is tightened, compression between the nut and the taper in the fitting body swages (crimps) the ferrule onto the tubing and provides a secure connection. When properly assembled, high-pressure fittings should have no gaps and little or no diameter change between the tubing and the fitting body (Fig. 3.8c). Such connections are referred to as zero-volume or zero-dead-volume connections. Nearly all manufacturers standardize on nuts with 10-32 threads and use the same ferrule taper for fittings used with 1/16-in o.d. tubing (different standards are used for metric sizes). This means that fittings from different manufacturers are nominally interchangeable. However, different manufacturers’ fittings may have different port depths, which means that for stainless-steel fittings, where the ferrule is tightly swaged onto the tubing, the ferrule setback from the end of the tube (Fig. 3.9a) may vary between manufacturers. For example, Rheodyne injection 102 EQUIPMENT (a) (b) ferrule setback leaks void volume Figure 3.9 Effect of ferrule setback on compression fitting assembly. (a) Tube end showing ferrule setback; (b) assembly with (left) too large of a ferrule setback, causing leaks, or (right) too small of a ferrule setback, creating a void volume. valves have noticeably deeper ports than most other fitting designs. Mismatching of the tube end and the fitting body after the fitting has been initially assembled can lead to problems, as illustrated in Figure 3.9b. For the example with the large ferrule setback, the tube end will reach the bottom of the fitting port before the ferrule contacts its mating taper (left-hand side of Fig. 3.9b). This will result in a fitting that leaks. Usually the ferrule can be forced to slide down the tubing as the nut is tightened; in this case the leak may stop and the fitting will be properly assembled. When the tube end is moved back to its original fitting body, however, the ferrule will contact the fitting taper before the tube end bottoms out in the fitting port (right-hand side of Fig. 3.9b). This will result in a leak-free connection that has a small void volume at the tube end, which can cause band spreading. To avoid these problems, it is a good idea to stay with a single manufacturer’s fittings when stainless-steel fittings are used. High-pressure fittings are made of PEEK as well as stainless steel. When the ferrule is made of PEEK (or some other polymer), it is not permanently swaged onto the tube end so that the ferrule can be slid easily into the proper position when a tube end is moved from one fitting to another. An added convenience that is popular in this fitting design is the use of a knurled PEEK nut (Fig. 3.8d) that can be finger-tightened instead of requiring a wrench. Finger-tightened PEEK fittings can be used for high-pressure fittings in conventional HPLC systems where pressures of 6000 psi are not exceeded. For higher pressure applications, stainless-steel fittings are required. If a PEEK fitting leaks, it is a good idea to turn off the pump, loosen the fitting, push the tubing to the bottom of the fitting port, and retighten the nut before turning the pump on again. This will reduce the risk of having the tubing slip during tightening, leaving a gap, as in Figure 3.9b. Nuts, ferrules, and fitting bodies can be made of PEEK or stainless steel. It is common to use PEEK nuts and ferrules in stainless stainless-steel fitting bodies, but stainless-steel nuts are rarely used with PEEK parts. 3.4 TUBING AND FITTINGS 103 frit frit out in in (a) (b) Figure 3.10 Specialty fittings. (a) In-line filter showing filter body, nut, and replaceable frit. (b) PEEK low-volume tee-mixer. Images courtesy of Upchurch Scientific, Inc., a unit of IDEX Corporation in the IDEX Health & Science Group. 3.4.2.3 Specialty Fittings Two modifications of standard high-pressure fittings provide special benefits: • in-line filters • low-volume mixers The in-line filter (Fig. 3.10a) is a modification of the standard union that is used to connect two pieces of tubing. A small-porosity frit (typically 0.5 μm porosity) is used in the in-line filter to remove unwanted debris from the fluid stream. It is a good practice to use an in-line filter just downstream from the autosampler on every HPLC system. This adds an insignificant amount of extra-column peak broadening, yet prevents particulate matter from the mobile phase, pump seals, valve rotors, or samples from reaching the column inlet frit. (Some pumps have built-in small-porosity filters at the pump outlet. These serve to trap particulate matter from pump seals or other upstream sources. In the absence of such filters, some users install an in-line filter between the pump and autosampler to prevent particulate matter from causing problems in the autosampler. The column frit can seldom be changed without damage to the column, whereas the in-line frit is designed for easy replacement. When the in-line filter becomes blocked, as signaled by an increase in system pressure, the frit is replaced and the HPLC system is back in service 104 EQUIPMENT with minimal downtime. In-line filters are available from many vendors in either stainless-steel or PEEK construction, and with various frit porosities. A low-volume mixer can be used to convert high-pressure-mixing pumps (Section 3.5.2) used for conventional applications to LC-MS-compatible pumps. Typical high-pressure mixers have volumes of 1 mL or more, whereas low-volume static mixers often contain <10 μL of volume. The example shown in Figure 3.10b is made of PEEK and contains a 10-μm porosity frit to aid mixing, yet has only 2.2 μL of swept volume. Other designs of low-volume mixers are also available. 3.5 PUMPING SYSTEMS Nearly all HPLC pumping systems in service today use some variation of the reciprocating-piston pump. Hydraulic-amplifier pumps are used primarily for col- umn packing and also may be used for preparative applications (Section 15.2). Syringe pumps are used as infusion pumps for tuning LC-MS systems, but not for high-pressure solvent delivery. Piston-diaphragm pumps that were once used for HPLC systems are no longer popular. Reciprocating-piston pumps have evolved over the years into reliable units that can provide hundreds of hours of trouble-free operation without maintenance. The precision and accuracy of the pump and its associated mobile-phase mixing system are keys to the success of HPLC as an analytical tool. Most HPLC pumping systems sold for routine analytical work are designed to work at pressures of up to 6000 psi (400 bar), but most workers operate such systems in the 2000 to 3000 psi (150–200 bar) region. HPLC systems that are promoted for fast analyses or high-pressure work, especially with sub-2-μmparticle columns, have higher upper-pressure limits and may allow routine operation in the 8000 to 15,000 psi (550–1000 bar) range or higher. It should be noted that conventional systems operated in the 5000 to 6000 psi region are able to provide some of the benefits of faster runs with smaller particle columns, without the need to purchase specialized equipment. However, when conventional HPLC systems are used at higher pressures (e.g., > 3000 psi), care must be taken to prevent leaks. For example, injection-valve rotors may need to be tensioned for higher pressures and fittings may need to be tightened more. Also, the mechanical wear rate of pump seals, injection rotor-seals, and other moving parts usually increases as the system operating pressure is increased. 3.5.1 Reciprocating-Piston Pumps The single-piston pump shown in Figure 3.11 is the core of all other HPLC pumping systems. The main components are: • motor •piston • pump seal • check valves The rotation of the pump motor (driving cam) drives the piston back and forth in the pump head. The piston usually is made of sapphire, although some pumps 3.5 PUMPING SYSTEMS 105 Intake Cycle outlet check valve piston driving cam (motor) inlet check valve Delivery Cycle pump seal piston (b) (a) Figure 3.11 Single-piston reciprocating pump. (a) Intake (fill) cycle, (b) delivery cycle. use ceramic pistons. A polymeric pump seal is used to prevent mobile phase from leaking out of the pump head. An inlet and outlet check valve control the direction of flow of the mobile phase. During the fill cycle (Fig. 3.11a), the piston is pulled out of the pump head, which creates a low-pressure region in the pump head. The outlet check valve closes and the inlet check valve opens, which allows the mobile phase to enter the pump. During the delivery cycle (Fig. 3.11b), the piston moves into the pump head, the inlet check valve closes, and the outlet check valve opens, which allows mobile phase to flow to the column. A dependable check valve is important for reliable pump operation. Ball-type check valves are commonly used, as illustrated in Figure 3.12a. The valve comprises a ruby ball and sapphire seat. This combination of materials gives a reliable seal, usually with no assistance other than gravity plus the pressure differential in the . the total of the autosampler -to- column plus column -to- detector connections. Sometimes it is advantageous to use a short piece of 0.007-in i.d. tubing between the autosampler and column, so as to. with pressures up to 6000 psi between the pump and detector, so the tubing must be able to withstand such pressures. Also tubing used to transport the sample from the autosampler to the column and. sizes from 0.005-in to 0.046-in i.d., whereas PEEK covers the 0.0025-in to 0.040-in range. The 0.010-in and 0.020-in i.d. sizes are used to transport solvent from the pump to the autosampler, where