1) Turn on the helium tank. Open the valve (red circle) all the way.

2) Turn on the deionized water valve (arrow). Open the valve all the way counterclockwise. The valve is made of Teflon, which is a very soft plastic. Do not force it.

3) The float (red circle) in the flow meter should rise to the top of the meter column.

4) Turn on the column oven and the conductivity detector switches (red circles). Normally the other modules are powered through the oven, unless someone has been fooling with the switches. At this stage, the pump, conductivity detector, eluant generator, oven, and autosampler (P, CD, EG, O, A) should all be on.

5) Check to make sure the column outflow tube is in the right place. The orange tube from the top of the column (vertical black tube) should go to the round 'T' block, as shown here. This tube may go to the Eluant in port on the suppressor (the light gray box behind the column). If you move the outflow tube, be sure you put a plug into the empty port.

6) On the front panel of the pump, press MENU, 8, 7. This lets you monitor the LEFT and RIGHT 'P-POINT' values when the pump is running.

7) Turn on the computer and the printer. On the computer, start the RUN program.

8) On the main menu bar of the RUN window choose FILE, SELECT METHOD. Choose the proper method (usually 'Silica in natural waters'), and agree to whatever the window asks. The pump should start. For this GP50 pump the LEFT and RIGHT P values should settle down to ~10 units difference within a few minutes. If it doesn't, you need to get the bubbles out of the pump head.

9) Get all of your samples ready. On the computer, start the SCHEDULE program.

10) The schedule is a list of all of your samples, plus calibration standards, plus a few other things. First should be three or so test samples. They can be a standard or unknown or a mix. Their purpose is to let the columns flush out and to make sure everything is working OK. Next should be the standards. Under "Sample Type" these are listed as calibration standards 1, 2, 3, and so on. The numbers refer back to the component table in the Method, which is identified in the widest column. Immediately after the standards should be deionized water, just to make sure the standards are washed out. Next are your unknowns, which can be assigned dilution factors if they have been diluted. Last are three samples to finish things up with: one standard to make sure everything went OK, and two of deionized water. The first deionized water should have the method: Flush silica.met. This turns off the PCR flow. The method for the second deionized water should be Stop silica.met. This method turns off the pump, detector lamp and other electronics at the end of the run. Save this schedule with the name 'Silica' followed by today's date (don't use slash marks).

11) The box thing on the left in this image is the post-column reagent (PCR) delivery system. It basically has a vessel pressurized with helium (red circle), inside of which is a bottle, like the one on the right side of the picture, that contains the post-column reagent. The helium pressure forces the PCR solution through a Teflon tube (arrow) into a mixing-T inside the oven. The PCR solution reacts with silica coming out of the column to form a yellow silico-molybdate compound, which is measured in the UV/visible detector at a wavelength of 410 nm. Vessel pressure is monitored with the gauge (yellow circle), controlled by the regulator (green circle), and the gas pressure is turned on and off using the gas valve (blue circle).

12) Look at the bottle in the pressure vessel. If it has enough solution leave it alone. You need about 10 ml per sample, and a full bottle can safely run 80 samples including standards. If there isn't enough, make more: Fill a 1000 ml plastic volumetric flask ~90% full with DI water. Add in the following order:

 

 

Gently mix the solution, trying not to make foam. Pour it gently into the 1 liter PCR bottle, put the bottle back in in the pressure vessel (red circle), and clamp the vessel lid back on. Turn the gas valve on (blue circle). Usually the best pressure is close to 40 psi.

13) The pressure in the pressure vessel forces the PCR into the eluant stream against the back pressure of the eluant flowing through all of the tubing from the mixing-T, through the reaction coil, through the detector, and to the drain. If the pressure vessel pressure is less than this back pressure, zero PCR flows and no colored silica complex is formed. The graph here shows how silica signal changes with PCR vessel pressure. In fact, the peak varies with tubing lengths and arrangements, and it is currently ~39 psi. Note, however, that the signal falls rapidly if pressure is too low, so it is better if pressure is a bit too high. You may wish to run some test solutions before your actual analytical run, at different PCR vessel pressures, to make sure the pressure is optimum.

14) Load your samples into the autosampler racks. The first sample should be right most next to the black dot (yellow circles). Sample order should be like that below. The numbers below refer to the row numbers in your schedule (see above), not any sample numbers you may have written on your unknowns. Notice that the last sample (deionized water) has its top pressed down only part way using the other end of the cap pusher.

 

15) Load the autosampler with the racks. The first rack should be in the front, dots to the right.

 

16) Press the Hold/Run button on the autosampler (red circle)so that the green light is lit on the Run side. This puts the autosampler under computer control.

17) In your sample racks in the autosampler you MUST put in two sample vials of DI water at the end. In the schedule, these two samples MUST run with the methods Flush Silica (second to last) and Stop Silica (last). Flush Silica closes the valve to the post-column reagent, sets the eluant to a low concentration, and runs the DI water sample the column. The purpose is to flush the post-column reagent out of the plumbing. Stop Silica turns off the pump and the detector lamp. This was described in item 5, above, but it is important to make sure you actually do it.

18) On the main menu bar of the RUN window choose FILE, SELECT SCHEDULE (circled in red). Choose the proper schedule (the one you just saved), and agree to whatever the window asks. Press the start button (circled in yellow) to start the run. Listen for the autosampler to start humming. Sample should inject at 2.30 minutes. Watch the chromatograph as it comes out on the screen. Keep an eye on the test samples to make sure the chromatograph looks good.