Overview TSip™ fitting developmentThe Tip Sealing approach is important to the attainment of the goals of UHPLC, namely,
- Reduced environmental impact due to reduced (perhaps 90% less) solvent useage.
- Reduced cost of analysis due to reduced solvent useage.
- Reduced cost of analysis due to reduced time of analysis.
- Increased usefullness of scientific data due to improvements in resolution.
- 1. The relationship between confidence in making a leak-free connection and acheiving a low-dispersion connection are very significantly improved over first generation UHPLC fittings. The strengthening of this correlation cannot be underestimated in it's ability to save time and effort in mis-directed troubleshooting. Previous (cone ferrule type) fittings could easily be installed in a leak-free manner, yet create adequate dead volumes to create loss of UHPLC levels of resolution. The impact of the false sense of confidence in attaining a 'good seal', and the unfortunate potential consequence of then proceeding to diagnose perceived faults of the column or instrument, cannot be ignored.
- 2. It is also significant that as compared to all 'coned ferrule' fittings, uncertainty about dispersion due to port machining variance is eliminated. That is, if one examines the CPI port, it can be seen that in order to seat tubing at the extreme bottom of the port, there must be at least enough clearance around the tubing to allow it to slide to the bottom of the port. Typical tubing is 1/16" of an inch in diameter, or 0.0625". CPI ports have been measured using precise pin guages, and found to vary in diameter from 0.064" to 0.067". This difference, between 0.0625 and 0.063" (best case) creates a hollow cylinder, which is a dead volume. The question might be, is this a swept or unswept dead volume? Better still, is this a 'partially swept' dead volume based on factors such as roughness at the port bottom? While it is possible to 'properly seat' tubing using a 'coned ferrule' (pre TSip™) fitting, it is difficult to predict exactly how the intrinsic cylindrical dead volume comes into play.
The approach of 'Tip Sealing' (TSip™) bypasses this cylindrical volume entierly.
This figure shows two effects which are inherent to the port. First, a sample (blue) in injected in the solvent stream. Both dilution and tailing are shown by how the (cyan) sample behaves. For good measure, slipping is shown. Finally the animation repeats.
The effect of tubing slipage on column resolution is shown here.
As instruments become more capable of UHPLC resolutions (whether or not operating at high pressures) it is of significant economic interest that uncertainties of 'connection' issues be minimized, since dispersion variance can lead to significant amounts of lost time in diagnosing system problems. Further, it seems intuitive that method validation protocols must take variability of extra-column dispersion into account, therefore, to reduce this cost, fittings which allow for greater possibility of extra-column dispersion variance should be minimized.
The so-called 'First Generation' UHPLC fittings have a common element, that is they use a cone-shaped ferrule to acheive a pressure seal. There are variations in the design of this ferrule, but with each design, the primary method to reduce dispersion is to attempt to 'bottom' the tubing in the port, thus minimizing swept dead volume. However; this first generation is subject to the uncertainty of 'indeterminate sweeping' of the cylindrical dead volume surrounding the tubing (at the port bottom).
It is speculated this thin cylindrical volume may not neccesarily be universally swept or unswept. Since no provision is made to attempt to seal the microscopically rough bottom of the port, pressure fluctuations in the flow path could conceivably communicate sample into this volume. This 'thin cylindrical volume' can easily be in the 400-600 nL range using measurements of typical ports. Some manufacturers appear to be aware of this possibility and minimize the diameter of the port such that it more closely matches that of the tubing. This 'tight fit' is desireable, but special care must be taken in design of TSip™ style fittings to avoid 'sticking' of the tip, as has been observed in early TSip™ prototypes.
Additional arguments for TSip™ styles are based on the number of potential re-uses of TSip™ fittings. The actual cost of fittings per instrument on a per-use basis now favors the TSip™ technology. This applies to all high resolution methods, those at high and low pressures. Lastly, the actual time and 'fussiness' required to make a consistent connection with 'First Generation' UHPLC fittings necessitates significant operator awareness and training on the use of the fittings themselves. The TSip™ style fittings finally return to the old and simple rule of merely "tighten a fitting until is does not leak, no more". This rule applied nicely 20 years ago, and applies again today. But through the era of 'First Generation' UHPLC fittings (cone ferrule types), the persistance of the rule (tighten till the leak just stops) may have been responsible for significant 'finger pointing' as UHPLC resolutions were improved. This 'finger pointing' refers to falsely blaming wither the column or instrument for inconsistency of performance which may have ben attributable to 'complex' fittings.
This series shows 4 variants of body styles, and the two major methods of Tip Sealing as claimed in US patent 9217522. They are not positioned chronologically, but the enumeration is chronological in terms of prototype fabrication and testing.
- Figure 4.1. 'Sliding Linking Tube' variety. This overall body style is described in US patent 9217522, but the features of the body style are not as yet claimed. This body style has a 'breakaway' feature described in the 9217522 patent, also unclaimed. At nearly 0.5" in diameter, this style can easily apply far more force than is necessary to seal using TSip™ (Tip Sealing).
What are generally claimed in 9217522 are two methods of Tip Sealig. The Tip Seal embodiment on this prototype is claim 1, which employs a slideable 'coupling tube' relative to the narrow diameter tubing. The seal in this case is simply a short section of 1/16" PEEK tubing, and is arranged in place of the malleable seal shown at the tip in the drawings of US patent 8006367
- Figure 4.2. This embodiment, third from left, was developed to test the 'fixed tube' method of the second claim of US patent 9217522, and is depicted in figure 1 above. Close examination will show a slight protrusion of the tip seal, coupled with internal (non-visible) geometry variations to test a slightly greater range of radial compression. Although the body style was completely acceptable for controlled evaluations, it was not suitable for general customer use. Additionally, the greater compression range of the protruded tip design allowed the seal to stick in the port, which was also found unnacceptable.
- Figure 4.3. This embodiment removes the hex portion of the body, favoring a relatively narrow body diameter (approx 6mm (0.25")) which limits the force a user can apply. This was found to be in the range of the optimum diameter for acheiving the required tightening force for 15,000 PSI seals. Although 20,000+ PSI is possible with slight additional tightening force, it is thought this is unnecessary and will lead to shortened life.
The tip and linking tube geometry are established in what is believed to be very close to optimum geometries, which acheive rapid installation and attainment of sealing forces, without sticking or other malfunctions.
- Figure 4.4. This embodiment is the current preferred body embodiment. Although it uses the 'hex-knurl' style, and it is conceivable an operator could over-tighten, the design does have a professional, compact, and functional form. The fine knurl, at 0.25" outer diameter acheives slightly higher tightening force than the design of figure 3, however an additional anodized aluminim labeling band, press fit to the body, does add a slight additional cost. This prototype series uses the optimized tip geometry of the series in Figure 3.