Persistent problems with high ECD background or "Ghost Peaks" during temperature program are almost always due to contamination introduced to the GC system from outside the GC, either from dirty samples, consumables, or from the carrier/make-up gas supplies.

INSURE CLEAN GAS SUPPLIES

Before beginning the troubleshooting process, make sure the GC system is installed with supply gasses of sufficient purity.

  1. The carrier and makeup gas purity should be >99.999%.
  2. After you are sure of the purity of the carrier gas, insure that the tank regulators have STAINLESS STEEL diaphragms, equivalent to Agilent Technologies P/N 8507-0407 (CGA 580 in USA).
  3. Install new 1/8" copper supply tubing - Agilent Technologies p/n 5180-4196. (Many times "clean" tubing from other vendors has caused high ECD background). At the same time install new traps in both the carrier and makeup supplies - moisture trap p/n5060-9084 closest to the tank, and indicating O2 trap p/n 3150-0528 closest to the GC. Be very careful to leak test the entire plumbing manifold.

ISOLATE PROBLEM TO CARRIER OR MAKEUP GAS SUPPLIES

Determine what components of the apparent ECD contamination are from the carrier vs.makeup systems in the GC. Typically, sharp, well resolved peaks that elute during a temperature program with no injection are from the carrier/inlet system. Broader "humps" in the baseline are usually from the makeup system. Overall high background (>500 Hz) can be contaminated gas from either the carrier or makeup gas supply or a contaminated ECD cell. To determine the source of contamination, perform the following procedure:

  1. Remove the column from the ECD, and inspect the installation of the makeup gas adapter in the ECD cell. It is very common for the adapter to be installed too low in the detector. This can be determined by measuring from the bottom of the 1/4" swagelok nut to the bottom of the hex of the makeup gas adapter. The measurement should be 19-20 mm. If it is >22mm, then the adapter was installed incorrectly. This is because of a ridge that is machined into the entrance of the ECD cell that prevents the adapter from easily seating all the way. Wiggling the adapter while installing will allow it to go all the way in. If the adapter is not all the way in it can leak during a temperature program and can adversely affect linearity for pesticide methods. Correct Installation of ECD makeup adapter
  2. Remove the makeup adapter from the ECD and then unscrew the tip and take out the Giga-bore liner. Inspect the makeup adapter body for graphite deposits. Remove all graphite deposits and clean it thoroughly with methanol. Soak the removable tip in methanol also.
  3. Re-assemble the adapter WITHOUT the Gigabore liner. Cap off the adapter using a 5 inch piece of narrow bore column, capped at one end with a new septum. The other end of the column should extend one mm past the tip of the adapter. When troubleshooting contamination problems, it is wise to use vespel column ferrules instead of graphite. The tip should be tightened just past finger tight with a clean pair of pliers and the whole assy. cleaned with methanol before installing.
  4. Re-install the capped off make-up adapter, making sure it is fully inserted into the ECD cell. Use a new 1/4" Vespel ferrule to allow for easy installation. Re-tighten after the ECD has been heated.
  5. Reset the makeup flow to the original setpoint and bakeout the detector at 350 degrees for 1 hour. During this detector bakeout you can put the inlet into split mode with 200-300 ml/min of split vent flow (gas saver off) and bakeout the injection port at 275 degrees and the column at its appropriate bakeout temperature.
  6. After the bakeout, run a series of blank runs using the customer method. If the baseline is acceptable - free of humps or peaks, and stays under 1000 hz. during the temperature program, then the detector and makeup side of the system is clean.
  7. Any unacceptable baseline problems could indicate contaminated makeup gas, EPC module, makeup adapter or ECD cell. These must be addressed before proceeding.

EVALUATE THE CARRIER SIDE

Once the ECD and the makeup system are determined to be clean, then the carrier, inlet and column should be evaluated.

  1. Reinstall the column in the makeup adapter by the same procedure used to install the piece of column used to cap the detector: - remove the makeup adapter from the ECD and remove the short piece of column. - install new ferrule on the column and trim the column end. - install the column in the makeup adapter such that 1-2 mm of column extends past the tip of the adapter. It is best not to use the Giga-bore liner. A better liner solution is the new mixing liner (P/N G2397-20540) from the uECD. It is backwards compatible to 5890 SER II ECD. Wipe-off the entire makeup adapter with methanol and install the adapter fully inserted into the ECD.
  2. Bakeout the entire system for another hour with the ECD at 350 degrees and the inlet in split mode at 275 degrees. The column should be baked at the manufacturer's recommended bakeout temperature.
  3. After bakeout, reload the customer's method and make a series of blank runs (no injection) to determine if the contamination problem persists. Again a single, well resolved peak could be due to EPC module "O"ring contamination mentioned previously. Address per service note.
  4. If the contamination persists then perform a complete injection port maintenance, including a thorough cleaning of the shell weldment, replacing gold seal and liner. Also install a known good 30 meter/320uM HP5 checkout column to rule out column contamination.
  5. Typically these peaks from the inlet side are due to contaminated carrier gas supply, EPC module, insert weldment, injection port/liner or column.