Bare Glass versus Deactivation
Almost no chemists used borosilicate capillary columns after Hewlett Packard showed they could be made from fused silica capillary. Fused silica is far more inert than borosilicate so phases bled less, more kinds of phases could be applied and analyte degradation was dramatically reduced. Most people who use press fits use only quartz because they don't need deactivation where borosilicate press fits do (and deactivation fails from day 1, much like new cars depreciate). Borosilicate connectors are also more prone to developing leaks in use (its a CTE thing).
Injection port liners are still almost all made of borosilicate, because nobody has figured out how to make them in quartz (until now).
It may seem obvious that more inert materials are better choices for surfaces exposed to the often miniscule quantities of analytes that we deal with in chromatography, but most of us have no clear concept of just how different these surfaces are, or just what constitutes “inertness”.
Inertness refers to both activity and reactivity when referring to surfaces, glasses in this case. A good approximation of relative activity is made through comparison of that surface concentration of silanol groups (Si-OH) on glasses, just as reactivity is proportional to the concentration of heavy metals (see graphs, below -- note ordinate scale is logarithmic).
As chemists, some of us bristle when a distinction is made between “fused quartz” and “fused silica”, but in the high-purity glass business, we refer to the glass made by CVD (from hydrolysis of silicon tetrachloride) as "fused silica" to distinguish it from mechanically sorted "quartz" sands that are melted or "fused". In CVD, distilled SiCl4 is hydrolyzed to SiO2 – distillation lowers the metals content but hydrolysis raises the silanol content and incomplete hydrolysis elevates the chlorine content far above that of natural or fused quartz.
As to the question, “should I used deactivation or not?” there is no clear answer except where it comes to borosilicate GC components, where the answer is a resounding “Yes!” Always deactivate borosilicate glass....except perhaps sample vials that are not heated.
As for fused quartz and fused silica, whether to deactivate is largely dependent upon analytes exposed surface area, residence time, concentration of analytes and the temperatures of those surfaces during analytes exposure. Again, note the ordinate is logarithmic in comparing exposed surface areas of various GC "inert path" components.
Deactivation of quartz and silica press fit connectors is pointless for most analyses because the worst case* exposed surface area is only about 60mm2 and residence time is brief, even if at the highest temperatures seen in the assay. Even so, where target analytes are both basic and trace quantity, deactivation is a warranted precaution.
The exposed surface area of injection port liners is significantly larger than that in press fit connectors and residence time can be much longer. Even if the temperature is somewhat lower, the primary reason that liners are always deactivated is that they are nearly always made of borosilicate glass.
Not all deactivations are created equal, of course, and not one of them is perfect. And 'silanol' comes in more forms that you might expect -- iso, gem, vic – and no deactivation covers them all. In fact, the best deactivations leave billions of gaps with exposed silanol or multiple silanol groups in various states of steric hindrance.
Imperfect to begin with, all deactivation coatings degrade with use, steadily increasing the exposed [silanol] (and [metals]), and some types work better with only certain classes of analytes. IQ recommends SilicoNert® 2000 deactivation (by Silcotec®, also known as Siltek®) for its broad compatibility, high thermal stability and durability.
SilicoNert 2000 on IQ’s fused quartz ZenLiner™ and Pres2Fit™ components is a low cost insurance policy against analytes loss; even as the deactivation degrades -- and it does degrade -- the level of underlying activity that is exposed is orders of magnitude lower than that of competing products.
* 0.53mm column to 0.53mm column connection -- but this area does have about 20 trillion silanols