In 1971, Feehs¹ was the first to recognize the UV transmitting properties of  Teflon®  tubing and to use it in a photochemical reactor.  This reactor synthesized methyl chloride from methane and chlorine gas by flowing the gases through the tube, irradiating the tube with UV light and thereby conducting a radiation-initiated chemical reaction. Although Iwaoka et al.², in 1976, were the first to use photohydrolytic detection in an HPLC post-column photochemical reactor, they used fused silica tubes in their apparatus.

      In 1980, Scholten et al³ published research on using  PTFE coils to substitute for quartz capillaries in a photochemical reactor.  Quartz capillaries have several disadvantages: they are not readily available in different geometries (coil and helix, diameter, length); they are expensive and fragile; and tight connections are difficult to make. Also, according to Bieler et al⁴, during photochemical reactions, hard-to-remove, UV-absorbent deposits, which are generally colored, commonly form on the surface (silica, usually) through which the radiation enters the fluid being treated.  However, according to Veloz⁵, a fluorocarbon UV-transmitting surface minimizes such deposition, and even if the capillaries eventually "gunk up", they are relatively inexpensive and can be discarded.

     Scholten used an air cooled (fan) 200 Watt  xenon-mercury lamp and observed comparable or better photolysis with the PTFE coils than with quartz.  He also obtained improved peaks with better symmetry and less tailing, which the different wetting characteristics of PTFE versus quartz may have caused.  The transparency of PTFE towards UV light may originate from a diffuse radiation transfer, and the high efficiency of the PTFE coils derives from the "light tube "effect.  This causes internal multiple reflectance of the radiation and hence amplifies the photochemical effect.  The light conducting properties of PTFE tubing are a well known undesirable phenomenon in fluorescence detection.  Scholten noted that this photoreactor performed best for a given analysis when he adjusted the length of tubing wound around the lamp and the flow rate to obtain the optimum residence time (exposure) to the UV light.  He also noted that aging of the light source and, the related temperature and UV characteristics of the light source influence the reactor performance considerably and should therefore be tested at regular intervals.

     Gandelman and Birks⁶  used a nitrogen-flushed photoreactor chamber to produce photoreduction with PTFE coils and an intense 254 nm. irradiation source (Pen Ray lamp, UV Products, Inc.).  The nitrogen prevented oxygen permeation through the PTFE tube.

     Ciccioli et al⁷ substituted Teflon FEP tubing for quartz in their photochemical reactor and found no major difference in the photochemical ionization efficiency between Teflon FEP and quartz capillaries.  They claim that Teflon FEP tubing (1/16 inch O.D., 0.015 inch I.D., calculated wall thickness 0.024 inch) "can easily stand pressures of the order of 5 kg/cm² ".

     Shih and Carr⁸ used a home-made photochemical reactor (medium pressure mercury lamp) with Teflon coils and air cooling (fan).  They mention that the cooling fan was necessary to prevent the  solvent from boiling at low flow rates, and that the reactor must reside in a hood because the mercury lamp generates ozone.  They also state that although temperature and oxygen content may be significant variables, they made no attempt to control their effects in this work.  They further note that controlling oxygen is difficult, especially with a photochemical reactor made of Teflon.  They found, as have other investigators, that short irradiation times result in insufficient yield of fluorescent product, and excessive irradiation times lead to a decreased fluorescence because of subsequent reactions.

     Engelhardt and Neue⁹ introduced the use of "deformed PTFE capillaries"  in a HPLC post-column non-UV reactor. They stated that an advantage of these tubes is that "very aggressive reagents like concentrated sulfuric acid can be used even at elevated temperatures".

     Selavka et al¹⁰ crocheted Teflon tubing for a HPLC post-column UV photochemical reactor, taking care to avoid kinking the tubing during the construction process.  This kinking would reduce the lifetime of the tubing by increasing its back pressure.  They also reported that, under certain conditions, organic solvents other than methanol, ethanol, or acetonitrile may cause swelling, leakage, or rupture of the Teflon tubing, especially in a knitted, open tubular (KOT) reactor, where pressure is elevated.  In addition, in a UV photochemical reactor, fluoride ion is liberated from the tubing during irradiation with UV light; this causes the tubing to turn brittle and eventually rupture after about 700 hours of use.  This effect increases if proper cooling is not provided for the lamp and the KOT assembly.

     Poulsen et al¹¹ reported that coils of PTFE tubing display good transparency to UV radiation in the wavelength region 230 - 400 nm.  While the oxygen permeability of PTFE can be advantageous in photo-oxidation reactions, the oxygen can interfere with other reactions.  In such a case, the reactor may reside in a nitrogen-flushed chamber.  When using sources that emit short wavelength radiation (less than 254 nm.), you should use thicker walled tubing because thin walled tubing tends to become brittle and burst. Batley¹⁶ has observed the release of fluoride ion and H+ ion from PTFE tubing irradiated at short wavelengths, which may explain the photo-brittling effect. Solvent composition may affect the magnitude of these releases allowing successful use of PTFE coils with a photoconductivity detector⁷. Researchers have observed PTFE photodegradation more severe in acetonitrile than in methanol.  The 200 Watt xenon-mercury arc lamp caused the coils to degrade slowly, even when methanol was the mobile phase polarity modifier¹²

     In 1993, Engelhardt et al¹³ used a low pressure mercury lamp because it worked without forming ozone while delivering UV radiation of 3.5 Watts.  They reported that Tefzel capillaries (20m x 0.3 mm I.D.) were superior to Teflon for photometric transparency (approximately 20% better) and long term stability.

     Lurie et al¹⁴ report that the two commercially purchased PTFE reactor coils (Aura Industries, Inc.) used in their study lasted for 150 and 80 injections, respectively, before developing leaks.  These investigators speculate that ice bath cooling for the lamp and the KOT assembly might alleviate this problem. In addition, they suggest that the back pressure generated by the detector (4.4 - 6.1 MPa) may contribute to coil rupture, since the PTFE tubing used was rated only for 3.4 MPa.

     Patel and Moye¹⁵ used a UV mercury lamp surrounded by an 0.8 mm. x 5.6 m knitted Teflon tubing as a photolytic reactor.  They found that the mobile phase used in the photolysis mechanism was important. Researchers have reported that the presence or absence of oxygen affects the rates of photolysis and the products formed from some pesticides.  Even though the solvents used in their research were purged with helium to avoid bubble formation during chromatography, the oxygen permeability of Teflon prevented deoxygenation from occurring during photolysis.  Therefore, we don't know the role of oxygen, if any, in the photolysis of these pesticides.

     Batley¹⁶ used an 8 foot length of Teflon tubing (1/16 inch O.D. x 0.012 inch I.D.) coiled around a PenRay lamp (UV Products, Inc.) enclosed in an aluminum block.  The temperature on the outside of the Teflon tubing reached 48 degrees C because cooling was not possible with this configuration.  However, in related experiments on Teflon bottles, Batley found that reducing the temperature from 55 degrees C to 30 degrees C with a cooling jacket decreased the rate of formation of  fluoride significantly. He used  water flowing through the coil and noted that fluoride release approximated a linear function of solution residence time in the coil (flow rates of 0.5 ml/min. and 1 ml/min.).

     Jardim et al¹⁷  have observed acid absorption and subsequent release in PTFE which was previously cleaned with warm concentrated nitric acid and subsequently used in contact with neutral or basic solutions. This drastically changes the pH of such a solution, and occurs without UV exposure.

     Conboy and Hotchkiss^{18 & 19}  investigated several sizes and types of Teflon [tetrafluoroethylene (TFE), perfluoroalkyl (PFA), and fluorinated ethylene polypropylene (FEP)] microbore tubing for irradiation coils with wall thicknesses of 0.15 - 0.41 mm.(0.006 - 0.016 inch).  They combined the coils with a 200 Watt mercury vapor lamp air cooled with a cooling fan.  The temperature of the air leaving the cooling chamber was 30 - 40 degrees C.  Acetonitrile served as an organic modifier with gradient elution.  Although the Teflon worked initially, Conboy and Hotchkiss noted a loss of response after only 2 to 3 weeks of use. They could not recover the loss by washing the tubing, and could only recover it by replacing the tubing with virgin material.  Conboy therefore converted his coils to thick-walled glass capillary tubing.  However, this glass tubing did require cleaning with chromic acid every 2 weeks when used for very dirty samples such as urine.

     Miles and Moye²⁰ noted that when large percentages (80 - 100%) of acetonitrile operated as solvent, the Teflon tubing developed several leaks and appeared to "sweat".  They avoided such solvents because they required replacing the photolysis coil.

     Crine²¹ tested Teflon FEP bottles filled with 22° C ultra pure water without UV irradiation and found that they released "significant amounts of fluoride ions "  into the water over a 30-day period.  Although he detected no other inorganic ions, the fluoride ion content was "so large that it affected the water conductivity."  There was an enormous increase in the rate of release and the amount of fluoride ions in bottles stored at 75°C, and no saturation occurred even after 60 days at 75°C, indicating that all ionizable impurities were still not extracted.  Since it is extremely unlikely that water ionizes free fluoride atoms, these ions could originate as propionate ions produced as oxidation by-products during the high-temperature manufacturing of the plastic.  These fluorocomplex ions would be detached from the main molecular chains, explaining why water can remove them easily.

     Werkhoven-Goewie et al¹² used two types of PTFE coils (with a 200 W xenon-mercury lamp) one with a "thin wall" [0.25 mm. (0.010 inch); 0.4 mm. I.D. (0.016 inch)] and one with a "thick wall" [0.65 mm. (0.026 inch); 0.3 mm. I.D. (0.012 inch)].  These investigators measured the fluorescence of phenol formed in the photoconversion of chlorinated phenols treated in their photoreactor.  As a mobile phase, they used pure methanol, water, methanol:water (65:35), hexane, dioxane, and tetrahydrofuran.  They could not use acetonitrile because, even with new coils, the PTFE capillary became porous immediately.  Although these investigators deoxygenated their mobile phase with nitrogen gas, they made no attempt to eliminate oxygen from the PTFE tubing environment. After 2 months of "continuous use",  the thin-walled capillary became porous and was replaced by another, taken from the same batch.  "For unexplained reasons," all the later capillaries became porous within several days.  They therefore continued their study with the thick-walled capillary, but found the optimum irradiation time was much longer now.  They could use the thick-walled coils for about 2 months before they became porous and had to be replaced.  Although coil-to-coil reproducibility was not thoroughly studied, they state that coil replacement does not lead to noticeably different results.  They recommend studying in more detail the influence of the type of PTFE coils used.

     Bachman et al²² state that high temperatures in the reaction coil can cause unwanted side reactions such as polymerization and that temperature fluctuations can decrease system reproducibility.  They also note that using high-pressure light sources (versus low-pressure mercury lights) can shorten the lifetime of PTFE irradiation coils which become brittle on prolonged exposure to high-intensity UV light.  Finally, both the intensity of the source and the length of the irradiation time influence the concentrations of intermediate short-lived radicals and excited molecules.

     Scholl et al²³ used a home-made photochemical reactor with a water-cooled compartment which kept the PTFE tubing at  5 -15 °C. They  found that the tubing lifetime was more than 20 weeks of operation and that the tubing seems not to be significantly affected by the irradiation (high-pressure mercury lamp) or the eluent (water: acetonitrile (70:30)).

     Dou and Krul²⁴ observed that using high-percentage organic solvents in the mobile phase except methanol, damaged the PTFE tubing. They could use 90% methanol in water in the mobile phase without dissolving the tubing.  High percentages of isopropanol in the mobile phase, however, could shorten the reactor's life.  Unfortunately, they could not use  acetonitrile: water in their case because acetonitrile dissolves the PTFE tube.  Isopropanol was compatible as long as the percentage did not exceed 30 %.

     Aichinger et al²⁵ used 12.5 m. of PTFE tubing with an 8 Watt "black lamp" and 80-95% acetonitrile as a mobile phase. They made no comments about the stability of the PTFE tube under these conditions.

     Kwakman et al²⁶ used a 90 Watt Philips mercury lamp. They  state that acetonitrile is not compatible with a PTFE coil in post-column photochemical reactors because it forms radicals and causes rapid leakage of the PTFE coil.

     Salamoun et al²⁷ used a PTFE capillary wound around a tubular 8 Watt low-pressure mercury lamp.  Because of the low power rating of the light source, the photoreactor required no active cooling. The mobile phase consisted of acetonitrile: water: tetrahydrofuran: glacial acetic acid (68:27:8:0.2 v/v).  The researchers made no comments regarding the stability of the PTFE tube.

     Mawatari et al²⁸ used a PTFE tube wound around a 15 Watt "black light" and methanol, ethanol, 1-propanol, 2-propanol, and acetonitrile as post-column reagents. They mention no PTFE tubing instability.

     C. de Ruiter et al²⁹ used a 90 W mercury lamp cooled by a fan and a 130 mm. long, 0.3 mm. i.d., 1/16 inch o.d. PTFE tube.  Both methanol and acetonitrile were used, however, irradiated acetonitrile: water mixtures lead to rapid destruction of the PTFE coil.

     In 1988 Lookabaugh and Krull³⁰ immersed their PTFE reactor coil in an ice water bath during irradiation with a mercury lamp while using a mobile phase of 5 mM tetrabutyl ammonium hydrogen sulfate, dissolved in a methanol : phosphate buffer (10:90), pH 6.8.

     In 1990, Dou and Krull³¹ also immersed their PTFE reactor coil in an ice water bath maintained at 0 - 5° C during irradiation with a low pressure mercury lamp and a mobile phase similar to Lookabaugh and Krull, above.