Optimising Manufacturer Injection Protocol

A recent addition to the laboratory here at Reach Separations is an Aurora Fusion SFC. Another analytical system that we hope to put into use to speed up screening, both for chiral and achiral separations. However, not being ones to accept straight off the shelf equipment, or out of the manual methods, part of the fun was having a play around with the injection protocol.

We wanted to optimise the injection protocol in order that we could obtain reproducible injections and maximise linearity of response. The Aurora Fusion utilises an injection protocol which draws a small defined amount of air both before and after picking up a user defined volume of sample. The Agilent recommended protocol picked up 1.5 uL of air before the sample and 5 uL of air post sample, making the order injected into the seat 5 uL of air, user defined volume of sample followed by 1.5 uL of air. The initial pick up of 1.5 uL of air is intended to push the full volume of sample out of the injection seat and into the loop.

Knowing that the volume of the injection seat on the system is 2.3 uL, we chose this value for the initial air pick up and matched this for the post sample pick up. We injected a series of volumes; 1 uL, 2 uL, 3 uL, 4 uL, 5 uL, 6 uL, 7 uL and 8 uL and measured the area of the peak response.

We plotted the results as shown as below:





This gave an R2 of 0.999 a good start for a linear response. However, two key points were noted; once above 6 uL the response began to plateau (data omitted from plot) and for an injection of 1 uL no response was seen.
The first point would be expected when using a 5 uL loop, once the loop is saturated, injecting more sample will simply push existing sample out of the back of the loop.

The second point, a lack of response from a 1 uL injection suggests that the sample is not reaching the loop – the push volume is not large enough. The correct push volume can therefore be calculated theoretically from the graph.

From the graph above x is calculated to be 1.53 adding this value to our 2.3 uL gives us 3.8 uL. A push volume of 3.8 uL was used and the same set of data collated and plotted.

In this instance, although x has decrease (to 0.71) the value of R2 has also slightly decreased. A push volume of 3.8 uL was tried again, this time including an equilibration time of six seconds.

The equilibration time allows for the re-expansion of any compressed air, ensuring the entire liquid sample is in the injection loop before the valve switches. This gave an R2 of 0.9989 and an x=0.42, an improvement. The 0.42 was added to the 3.8 uL and the experiment repeated, again using the equilibration time. This procedure was then repeated and the push volume increased until no (or only marginal improvements) improvements were seen.

The data below details the R2 value and x for each of the experiments:


The push volume we have selected for our injection protocol is 4.2 uL with a six second equilibration time. With the closed loop system that is on the Aurora fusion we have accepted that at very small volumes, <1.5 uL, we will not be able produce a linear response, realistically we have a system that will give a linear response from 2-5 uL.


Return to blogs

On the recruitment trail again. #NTUFair https://t.co/WPymj9qGmc