Sciencemadness Discussion Board
Not logged in [Login - Register]
Go To Bottom

Printable Version  
Author: Subject: HP (Agilent) 1050 - HPLC
DavidJR
National Hazard
****




Posts: 433
Registered: 1-1-2018
Location: Scotland
Member Is Offline

Mood: Anxious

[*] posted on 20-6-2018 at 12:54
HP (Agilent) 1050 - HPLC


So, I scored an eBay bargain: an old HP 1050 HPLC setup, for 99 pence. I did end up having to pay a courier £114 to get it to me, but even at £114.99 it's a bargain.


The biggest issue is that I don't actually have a column yet. I'll need to get one before I can do any chromatography, obviously. I'm thinking I'll buy a second-hand C18 (octadecylsilane/ODS) column but honestly I don't really know what I'm doing.

In the meantime I have been testing all of the modules individually and together as a system. I've been surprised by how little work it took to get it going.

I've taken all of the modules to bits, cleaned them out, and fixed various minor problems. Some of the modules had clearly been cannibalised for spare parts (and I took even more off of them) but with a small amount of effort I now have a full working set:

  • 1x quaternary pump
  • 1x variable wavelength UV absorbance detector
  • 1x autosampler
  • 1x solvent cabinet w/ column oven



The second quaternary pump and second autosampler have been broken down into the useful spare parts (which I'll keep and/or sell on eBay - but I'm going to throw out the bulky cases to save space).

The second variable wavelength UV detector complains with a 'low intensity' message, so I'm guessing that the deuterium lamp is near the end of its life, or perhaps there's an alignment issue in the optical block. Superficially it does appear to work though, but I haven't spent any time looking at it since the other one works perfectly.

The online vacuum degasser is not currently working. It runs for a bit but shuts off with an error light - apparently this means that the pressure didn't reach the threshold before a timeout, and it gave up. However the vacuum pump definitely works, so I think there is a small leak somewhere.

image-2.jpg - 1.6MB




View user's profile View All Posts By User
stamasd
Harmless
*




Posts: 43
Registered: 24-5-2018
Location: in the crosshairs
Member Is Offline

Mood: moody

[*] posted on 21-6-2018 at 00:54


This is the stuff of dreams. Or nightmares depending on the person.
I have spent many sleepless noghts back in the day when I was a graduate student baby-sitting a chromatography system.
I won't do it again.
View user's profile View All Posts By User
DavidJR
National Hazard
****




Posts: 433
Registered: 1-1-2018
Location: Scotland
Member Is Offline

Mood: Anxious

[*] posted on 17-7-2018 at 00:00


Some updates:

I fixed the degasser. Essentially it's just some gas-permeable tubing (not sure what it's made of, I think either PE or some kind of fluoropolymer) bundled inside a vacuum chamber. Anyway, the seals on the vacuum chamber weren't in great condition, so with the addition of some silicone gasket compound, it now works.

Bought a load of stuff:


  • 4 columns, all ODS / C18, in various sizes
  • misc PEEK and PTFE tubing/fittings (both of which are remarkably expensive!)
  • a pile of 0.45um PTFE syringe filters for filtering samples
  • some 0.5 micron inline filters (phenomenix "krudkatcher". yes, they actually named the product krudkatcher)
  • some autosampler vials
  • a few books on HPLC
  • appropriate solvents
  • a motorised rheodyne injection valve to replace the pneumatic one in the autosampler (since I don't have a compressor, I was abusing my poor vacuum pump's exhaust). This required a little bit of hackery/glue-logic to operate from the solenoid signals output by the autosampler's computer, and also to convince the computer that all is okay. It works very nicely though.


The oscilloscope was useful for initial tests but is not practical for actual use. Although it can save traces to a USB disk for later analysis, the software arbitrarily limits the maximum acquisition length to 10 minutes (and runs are often going to be longer than that). It is kinda annoying because it's not a memory issue, but it refuses to sample any slower than 20ks/s which is overkill. Also, like most DSOs the ADC resolution is only 8 bits, which isn't ideal for this purpose. Yet the temporal resolution is crazy overkill for chromatography (1Gs/s max) .

So: I have constructed a simple data acquisition device with an ADS1232 ADC (24 bits, however since I'm feeding a single-ended signal into a differential input, I only use half of the dynamic range). It can record to an SD card or directly to the PC in real time. It is also connected to the HPSystem remote bus to automatically start acquisition at the correct time etc.


[Edited on 17-7-2018 by DavidJR]




View user's profile View All Posts By User
brubei
Hazard to Self
**




Posts: 97
Registered: 8-3-2015
Location: France
Member Is Offline

Mood: No Mood

[*] posted on 17-7-2018 at 02:01


you can try this column https://labassets.com/auctions/product/33195
Quote: Originally posted by stamasd  
I won't do it again.

View user's profile View All Posts By User
DavidJR
National Hazard
****




Posts: 433
Registered: 1-1-2018
Location: Scotland
Member Is Offline

Mood: Anxious

[*] posted on 25-7-2018 at 06:26


My biggest problem so far has been water quality. Turns out distilled water really isn't good enough - at least, for gradient runs.

With gradient runs (starting with a weak eluent and gradually increasing the eluent strength) contaminants in the weak eluent (which for reverse phase is water) are retained strongly by the column and then elute later with the stronger eluent, creating ghost peaks.


Here's a blank run (i.e. no sample was injected) using distilled water:
gradblank.png - 27kB
Scale: 1 horizontal division = 5 mins, 1 vertical division = 40 mAU
Detection: UV absorbance at 254nm
Column: C18/ODS 50mmx4.6mm
Eluent: Hold at 20:80 (vol) MeOH:dH2O for 2 minutes, then linear gradient to 100% MeOH over 10 minutes, then hold for 3 minutes




Isocratic runs (i.e. runs with constant eluent composition) are much less sensitive to water quality issues because strongly retained contaminants are, well, strongly retained (may never elute at all or elute much later than all sample components). Of course, the accumulation of contaminants will gradually change the chromatographic properties of the column, but washing with strong eluent periodically will sort that.

Here's an isocratic blank run using distilled water:
isoblank.png - 8kB
Scale: 1 horizontal division = 5 mins, 1 vertical division = 40 mAU
Detection: UV absorbance at 254nm
Column: C18/ODS 50mmx4.6mm
Eluent: Isocratic, 66:34 (vol) MeOH:dH2O




View user's profile View All Posts By User
wg48temp
Unregistered




Posts: N/A
Registered: N/A
Member Is Offline


[*] posted on 25-7-2018 at 13:59


Quote: Originally posted by DavidJR  
Some updates:
So: I have constructed a simple data acquisition device with an ADS1232 ADC (24 bits, however since I'm feeding a single-ended signal into a differential input, I only use half of the dynamic range). It can record to an SD card or directly to the PC in real time. It is also connected to the HPSystem remote bus to automatically start acquisition at the correct time etc.
[Edited on 17-7-2018 by DavidJR]


I recognised that chip. 23.5 usable bits for about £5. Did use the evaluation board for it or did you design your own? How do you get the info in to the PC. Are you going to write up the details? Please do.

I recognised that chip. 23.5 usable bits for about £5. Did you use the evaluation board for it or did you design your own? How do you get the info in to the PC. Are you going to write up the details? Please do.

Great project by the way. It must have been a huge task getting it going and understanding it. Any particular plans on what you will do with it?
DavidJR
National Hazard
****




Posts: 433
Registered: 1-1-2018
Location: Scotland
Member Is Offline

Mood: Anxious

[*] posted on 26-7-2018 at 02:27


Quote: Originally posted by wg48temp  

I recognised that chip. 23.5 usable bits for about £5. Did you use the evaluation board for it or did you design your own? How do you get the info in to the PC. Are you going to write up the details? Please do.

Great project by the way. It must have been a huge task getting it going and understanding it. Any particular plans on what you will do with it?


I didn't use the evaluation board - just wired it up to an arduino I had lying around because I wanted to get something up and running quickly. The data is formatted as CSV, either written to the SD card or transmitted in real time over the serial bus. I will document the system a bit more thoroughly at some point but at the moment I'm still making changes to things so I'll hold off for now.

No particular plans for the HPLC - just for fun!




View user's profile View All Posts By User
DavidJR
National Hazard
****




Posts: 433
Registered: 1-1-2018
Location: Scotland
Member Is Offline

Mood: Anxious

[*] posted on 19-9-2018 at 08:41


I'm no longer using the DIY DAQ device as I've managed to obtain a copy of HP Chemstation.

I have bought a replacement install CD for version A.10.01, which is one of the last versions which supports the 1050, as well as the first version of chemstation which will run on Windows XP. I've set this up on an old pc, fitted with a National Instruments PCI-GPIB card. I have it working with the variable wavelength detector, which means I can record chromatogram data and process it directly in chemstation.

I haven't been able to get it to talk to the autosampler or quaternary pump. This means that I still have to enter numerous parameters on the instrument front panel manually, which is potentially error-prone; that the configuration is not automatically recorded with the run data for future reference; and that I can't really use the automation features to run sequences unattended. The problem seems to be down to an incompatibility with the interface layer which allows using an NI brand GPIB card with HP software. After a lot of faffing about with different driver versions etc, I have given up and ordered an old HP GPIB card, which should work, but it hasn't arrived yet.




View user's profile View All Posts By User
DavidJR
National Hazard
****




Posts: 433
Registered: 1-1-2018
Location: Scotland
Member Is Offline

Mood: Anxious

[*] posted on 22-9-2018 at 15:15


The HP 82350 GPIB/PCI card arrived today - I now have the chemstation software fully working.
View user's profile View All Posts By User
DavidJR
National Hazard
****




Posts: 433
Registered: 1-1-2018
Location: Scotland
Member Is Offline

Mood: Anxious

[*] posted on 10-10-2018 at 04:55


Main ongoing issue was enormous ghost peaks from mobile phase impurities. My own attempts at purifying water and methanol to a sufficient degree were not all that successful. To deal with this I have:

  1. Bought an old Milli-Q Synthesis A10 ultrapure water machine, along with a set of new old-stock consumables for it (they are very pricey).

    This takes in purified water (I'm feeding mine from a reservoir filled manually from an el-cheapo ebay reverse osmosis / mixed-bed DI system) and puts out Type I ultrapure water (resistivity = 18.2 megaohm centimetres at 25C, total organic carbon around 7 ppb).


  2. Bought HPLC grade solvents. These have been purified to remove UV-absorbing trace organic contaminants.

    I purchased HPLC grade methanol and HPLC Sp. grade acetonitrile from a local company, Rathburn Chemicals.


Together, these result in a dramatically cleaner baseline in gradient runs.

Here's two overlaid chromatograms, both no-injection blank runs of the same gradient method:

unknown.png - 26kB

Guess which one used AR grade methanol + triple-distilled water, and which one used HPLC-grade methanol + Milli-Q water...
View user's profile View All Posts By User
unionised
International Hazard
*****




Posts: 3831
Registered: 1-11-2003
Location: UK
Member Is Offline

Mood: No Mood

[*] posted on 10-10-2018 at 10:20


I always wondered what would happen if you tried vodka as a mobile phase...
View user's profile View All Posts By User
DavidJR
National Hazard
****




Posts: 433
Registered: 1-1-2018
Location: Scotland
Member Is Offline

Mood: Anxious

[*] posted on 10-10-2018 at 12:30


Quote: Originally posted by unionised  
I always wondered what would happen if you tried vodka as a mobile phase...


Hah! I imagine if you were to use vodka for isocratic runs it'd probably be just fine. Gradient runs, expect a crapton of ghost peaks.
View user's profile View All Posts By User
streety
Hazard to Others
***




Posts: 100
Registered: 14-5-2018
Member Is Offline


[*] posted on 10-10-2018 at 12:45


I remember seeing an article on this. Took a bit of searching but here it is: https://pubs.acs.org/doi/abs/10.1021/acssuschemeng.5b00133
View user's profile View All Posts By User
unionised
International Hazard
*****




Posts: 3831
Registered: 1-11-2003
Location: UK
Member Is Offline

Mood: No Mood

[*] posted on 10-10-2018 at 12:55


Thanks.
I was more or less kidding but...
View user's profile View All Posts By User
phlogiston
International Hazard
*****




Posts: 1202
Registered: 26-4-2008
Location: Neon Thorium Erbium Lanthanum Neodymium Sulphur
Member Is Offline

Mood: pyrophoric

[*] posted on 11-10-2018 at 07:13


Before you had the HP software, how were you programming the pump? Does it let you program gradients manually?



-----
"If a rocket goes up, who cares where it comes down, that's not my concern said Wernher von Braun" - Tom Lehrer
View user's profile View All Posts By User
DavidJR
National Hazard
****




Posts: 433
Registered: 1-1-2018
Location: Scotland
Member Is Offline

Mood: Anxious

[*] posted on 13-10-2018 at 11:17


Quote: Originally posted by phlogiston  
Before you had the HP software, how were you programming the pump? Does it let you program gradients manually?


Yes, you can program all of the functions via the front panel buttons. It appears that the GPIB card inside each module which is used for comms with the PC was actually an optional extra - so these were all designed to be usable without Chemstation.

--

Here's my current HPLC workbench (as usual, please excuse the mess):
image0-2.jpg - 2MB

[Edited on 13-10-2018 by DavidJR]
View user's profile View All Posts By User

  Go To Top