[RID Setting] How does auxiliary range effect HPLC analysis?

[Purplish3979]

Hello all,

Could someone please tell me what is the function of the 'auxiliary range' output setting of the detector (RID) and how does it effect the analysis? I came across this setting in the detector tab of the settings under the output section. Does it increase or decrease the speed of detection?

Please find the specifications of the HPLC below:
- Shimadzu LC20A
- RID: 20-A***
- RHM Column (300 x 7.8 mm)
- Matrix: Sulfonated Styrene Divinyl Benzene***
- Particle Size: 8 um***
- Size Exclusion (sugars); Ion Exchange (org. acids);Reverse Phase (alcohols)
- pH Range: Neutral***
- Extra column volume is large: atleast 60 cm long tubing of 0.3mm

- Mobile Phase: H2SO4 (15mM)***
- Flow rate:: 0.6 mL/min (Isocratic)
- Column Temperature: 60 degrees C***
- Cell Temperature: 60 degrees C***
- Method sampling rate: 25 Hz***
- Detector sampling rate: 50Hz (fastest possible)***
- Detector response time: 0.05 seconds (fastest possible)***
- -Recorder Range: 100 uRIU/FS

- Analyzing: Monosaccharides, organic acids and sugars***
- Internal Standard (acids): 200 uL of Crotonic acid (1.5 g/L)***
- Internal Standard (alcohol): 200 uL of Ethylene glycol (1 g/L)***

Important
- Coloumn: 1 year old and in good condition***
- HPLC: 13 years old and in bad condition***
- Computer: 20+ years old (windows XP)***

Notes:
***Added after edit on 14 August

[Multidimensional]

***This has nothing to do with the sampling rate of the detector.
The AUX OUTPUT selector you refer to changes the signal output of the voltage scale. My boss tells me, that this feature comes from the 1970's when it was common to need to match a detector's voltage output to the input voltage of the device that you used to record it with. Often, strip chart recorders were used and they used a scale of 0 - 10mVs for one arbitrary scale unit. So to match the detector output to the chart recorded, you would set it to 10mVs. For other systems such as an integrator (late 1970's / early 1980's) you would set it to 1000 mVs. When computers arrived, it could be set at 1 or 2 V full scale so you needed to review the details before selecting a value. Setting the value to the wrong # may multiply or divide the ACTUAL signal voltage measured so it is important to get it right (and always use the same settings).
* For a modern RI detector, it will usually be 0 - 1V scale (0 - 1000 mVs). For RI and UV/VIS detectors, the voltage units may be equated to arbitrary units, such as 'relative' RI units or milli-AU units per voltage range.

BTW: This should all be in the manual for YOUR specific detector and YOUR specific A/D module (CDS). They will tell you which scales to use for their own brand of instruments.

[Purplish3979]

***This has nothing to do with the sampling rate of the detector.
The AUX OUTPUT selector you refer to changes the signal output of the voltage scale. My boss tells me, that this feature comes from the 1970's when it was common to need to match a detector's voltage output to the input voltage of the device that you used to record it with. Often, strip chart recorders were used and they used a scale of 0 - 10mVs for one arbitrary scale unit. So to match the detector output to the chart recorded, you would set it to 10mVs. For other systems such as an integrator (late 1970's / early 1980's) you would set it to 1000 mVs. When computers arrived, it could be set at 1 or 2 V full scale so you needed to review the details before selecting a value. Setting the value to the wrong # may multiply or divide the ACTUAL signal voltage measured so it is important to get it right (and always use the same settings).
* For a modern RI detector, it will usually be 0 - 1V scale (0 - 1000 mVs). For RI and UV/VIS detectors, the voltage units may be equated to arbitrary units, such as 'relative' RI units or milli-AU units per voltage range.

BTW: This should all be in the manual for YOUR specific detector and YOUR specific A/D module (CDS). They will tell you which scales to use for their own brand of instruments.

Thank you so much for the explanation. Unfortunately, I've only just joined the lab where I'm working on the HPLC and they seemed to have lost the manuals. That means I'm pretty much on my own when it comes to fixing the HPLC, as others in the lab have limited experience with the HPLC and I couldn't find the manual for that specific RID on the internet either. (I have contacted Shimadzu though).

On an unrelated topic, I'm also trying to understand how the 'delay time' setting (currently set to 2 seconds) for the detector effects the analysis. Do you have any idea regarding this? The biggest problem I'm facing is that the detectors are unable to "complete" the peaks on time. The peaks rise fast enough but the drop from the peak top to the baseline is waaaayyyy slow. None of the settings seem to have any effect on this.

For example, regardless of the settings used the glucose peak will slow down once it reaches graphs the peak top. At 10.5 min RT, the chromatograph will ALWAYS stop graphing the glucose peak and start graphing the next one. The sampling rates, temperature, flow rate, response time and auxiliary range (as you explained) have no effect on this. So, I wonder if this could be resolved by changing the 'Delay Time' setting from 2 seconds to something else.

[uzman]

Are you sure that your column is in good condition ?
Too much tailing mostly related to column .

Also you may check column connections if there is a leak or dead volume.

Another possibility is a leaking valve ; which controls the flow to reference and sample cells.

[Purplish3979]

Are you sure that your column is in good condition ?
Too much tailing mostly related to column .

Also you may check column connections if there is a leak or dead volume.

Another possibility is a leaking valve ; which controls the flow to reference and sample cells.

The column is in terrible shape** (please see correction), and hasn't been properly maintained in years so I'm fighting an uphill battle. THAT SAID, this is not an issue of peak tailing. The best way to describe the situation is as follows:
- The peak rise/response time (time from 10% to 90% of peak) is very good, BUT
- The "peak fall" or decay time (from 90% to 10%) is slow

So, though the peak shape is great the chromatograph starts graphing the next analyte before it has finished graphing the peak of the first one. The peak fall is just too slow for some reason.

Edit:
I checked the column after reading the posts in this thread and it turns out the column is in very good condition.

[TylerSmith123]

Hi Purplish,

You very accurately described peak tailing right there. However, if you do insist this is not some sort of peak tailing you could always upload a chromatogram to give us a better idea of what you're talking about. Uzman may be completely correct about your column and it simply needs a replacement. Have you attempted to restore it in any shape or form? What are you even running and on what column? Some analytes can behave completely differently depending on the stationary and mobile phases.

[Purplish3979]

Hi Purplish,

You very accurately described peak tailing right there. However, if you do insist this is not some sort of peak tailing you could always upload a chromatogram to give us a better idea of what you're talking about. Uzman may be completely correct about your column and it simply needs a replacement. Have you attempted to restore it in any shape or form? What are you even running and on what column? Some analytes can behave completely differently depending on the stationary and mobile phases.

To clarify the problem, I think the column is resolving the peaks but the detector is too old/slow (13 years old) and is unable to detect them as separate peaks. Additionally, the computer and software are also over 20 years old. They simply cannot keep up. I want to make the distinction that there definitely is some peak tailing but that is a separate issue.

I checked the column again after reading your post and it turns out that the column is in very good condition and was installed only a year ago. There also doesn't seem to be any leaks. As for the steps I've taken to restore the column: washing (with water), recharging the column with H2SO4 and purging the line with IPA/Water. Unfortunately, I'm a lowly Masters student, so I'm not allowed to make any changes to the hardware.

I have added more information regard the machine in my first post, and will try to add the chromatgraph later. The issue I've described is for the first two peaks (glucose and malic acid) which are 30-45 seconds apart. Glucose is separated by size-exclusion whereas malic acid is by ion exchange).

[uzman]

Your explanation about the detector is not logical.
The detector sees what the column sends.
Shimadzu RID-20A is not an old detector , even the 30 years old RID-6A is working perfectly.

The detector response time may effect the integration , but not the peak shape.

By the way , did you flush the detector reference flow cell with fresh mobile phase , before starting the analysis ?

[DR]

Note - Google will find a surprising number of manuals for nearly anything and should be tried after searching the mfg website for anything they built this century. The manuals are usually out there, if you look.

[Purplish3979]

Your explanation about the detector is not logical.
The detector sees what the column sends.
Shimadzu RID-20A is not an old detector , even the 30 years old RID-6A is working perfectly.

The detector response time may effect the integration , but not the peak shape.

By the way , did you flush the detector reference flow cell with fresh mobile phase , before starting the analysis ?

Yes, I flushed the detector reference cell before starting the analysis. I also checked the analysis again today (at different settings) at it seems that:
- There is some peak tailing for glucose (first peak detected) but it was only noticeable after changing the temperature
- The "peak interference" I'm observing is not from the chemical I thought it was. I tried analyzing a new standard today and found the same peak shape. So, either that peak is from a contamination from some unknown source, or it's peak shouldering for glucose. I'm not sure if the latter makes sense.

Could you please explain how the detector response time may effect the integration? Maybe that's the cause of the problem and I've misunderstood what (and where) the problem actually was.

[TylerSmith123]

Hi Purplish,

Let me get this straight-- you injected a standard and you received the same peak shape for the standard analyte, or you injected your standard analyte and there was another peak that exhibited this peak tailing phenomenon? If so, then that would suggest contamination somewhere on your instrument but could be confirmed by running a wash sequence with a short EQ time and then another wash sequence with a significantly longer EQ time and then compared the contamination peaks to each other. More contaminant should build up the longer you're running a weak MP, while in the longer EQ this will exhibit itself as a larger peak. However, what is the retention factor of your glucose? Could it be eluting at the dead-time and this is why you're seeing these issues?