dodecane inconsistent in each injection

[tmalew2]

I am running dodecane pyrolysis and oxidation experiments and whenever I run an experiment at a temperature around 940K, the reaction time is around 2.3ms (dodecane doesn't react) and analyze the mixture the GC shows different levels of dodecane every time.

I was thinking maybe the fuel is condensing in the apparatus (shock tube) or the mixture tank but the shock tube is heated to 100 deg C and the mixture tank is at 140 deg C. There is about 54ppm of fuel in a 21 bar mixture.

Would the fuel condense in the tank? The partial pressure of the fuel in the mixture tank (21 bar) was around 0.78 torr. At 140 deg C and 1 atm, the vapor pressure is almost 109 torr and at 100 deg C and 1 atm the vapor pressure is around 28 torr.

Would the dodecane be condensing in the mixture tank if the tank is at 21 bar instead of 1 bar.

I am measuring dodecane on a Agilent 6890 GC using a PLOT-Q column (30m long, 0.32 um in dia.) all the lines leading from the shock tube
to the GC are at 150 deg. C.

I know its something with the fuel being heavy, as I didn't have any problems when running decane at the same conditions accept the mixture tank was at 100 deg C.

The mixture tank is made of carbon steel and its 42.47 liters. I also noticed larger discrepancies when I dropped the ppm levels from 75 to 50ppm even when the mixture tank was still 100 deg C. When I started to increase the temp. of the mixture tank the discrepancies of measuring the dodecane when no reactions occurs became larger.

Any thoughts?

Does anyone where I can find or how to find the vapor pressure of dodecane at elevated pressures and/or temperatures 10-50 bar?

[Peter Apps]

Condensation, in the sense of liquid coming out of the vapour phase, is probably not the problem. What is more likely is that you have adsorption onto surfaces. Increasing the temperatures of everything that the sample touches will help. Check out Tobias et al Journal of Chromatography A, 1140 (2007) 1–12.

Peter

[tmalew2]

Peter,

so increasing the temperature of the shock tube (experimental apparatus) to 120 deg C from 100 would help?

Wouldn't increasing the temperature increase adsorption as more sites open up as surface is heated.
Can the variation of temperature between the mixture tank (140 deg C) and the Shock Tube (100 deg C) be a problem.

[Peter Apps]

As a general rule; things stick less to hotter surfaces, so the hotter the apparatus the less adsorption you will get. Aim to have an increase in temperature from one stage to the next so that nothing can adsorb due to a temperature drop. I mixed up the references - you need Modey and Doskey, Journal of Chromatography A, 1121 (2006) 16–22. They were working with heavier compounds than C12, but had treated surfaces and much higher temperatures also - and they still found measurable losses to adsorption:

Abstract
Semivolatile organic compounds (SVOCs) are distributed in the atmosphere between the gas- and aerosol-phases. The low vapor pressures of some SVOCs makes thermal extraction and transfer through gas chromatographic (GC) systems difficult.We evaluated a programmable temperature vaporization (PTV)GCinlet, which served as the preconcentration module, and four open-tubular capillaries (Silcosteel- and Siltek-treated stainless steel, Silcosteel-treated stainless steel coated with 100% dimethylpolysiloxane, and deactivated fused silica) as transfer lines in a valveless, wholesample analytic system. Thermal extraction of C9–C36 n-alkanes at 300 and 320 ◦C from fused silica and quartz wool in the PTV inlet was equally efficient. Adsorptive losses of C22–C36 n-alkanes to stainless steel surfaces that protruded into the PTV inlet were suspected. Thus, treatment of the outer surfaces of transfer lines is recommended for effective thermal transfer of SVOCs. Transfer efficiencies began to decline after n-C24, n-C28, and n-C30 in Silcosteel-treated stainless steel, deactivated fused silica, and Siltek-treated stainless steel transfer lines, respectively. Thus, quantitative recovery at 320 ◦C of compounds with vapor pressures less than about 3×10−8 Pa is not expected in valveless SVOC thermal desorption systems
that use Siltek-treated stainless steel transfer lines and fused silica or quartz wool as preconcentration substrates.
© 2006 Elsevier B.V. All rights reserved.

Peter

[tmalew2]

Peter,

our transfer lines from the Shock Tube to the GC gas inlets are treated, they are either Siltek or treated deactivated fused silica. We measure really heavy species with our system, even 4 ring species when we run experiments. The problem I am having is the inconsistencies in the non-reactive range of dodecane (just around 50ppm). Could it be the carbon steel mixture tank we keep the dodecane in Argon mix. Could the carbon steel be causing the adsorption?