Mixing liquids onto other liquids is a very common step used in automated liquid handling workflows. Scientists often want to mix multiple liquids onto liquids that are already in the wells of a destination plate in a particular order (e.g., transfering mastermixes onto samples).

This technical note outlines a workflow that transferred liquids from a source plate and mixed them onto liquids present on another plate using the Mix Onto element. The workflow was executed on three liquid handlers: a Gilson PIPETMAX 268, a Hamilton Microlab STAR and a Tecan Freedom EVO.

This workflow has been used to validate liquid handling accuracy for the volumes of liquid (tartrazine or water) transferred onto liquids already in the wells of a plate to make mixtures, and provides a starting point for liquid handling optimization should it be required.

Within this workflow (Figure 1), four mixtures of 0.05 mM and 0.025 mM tartrazine were made using 2 distinct mixing orders; the first two mixing reverse osmosis (RO) water onto tartrazine, the second mixing tartrazine onto RO water.

Figure 1: Mix Onto workflow in Synthace. This workflow describes a method for using the Mix Onto element. The Define Liquids And Plates element defines the reagents to be used in the experiment. The first Aliquot element defines transfers of tartrazine or water at different volumes to the first assay plate, whilst the second Aliquot repeats the steps to a second assay plate. The Append element joins the two liquid sets produced by the two Aliquot elements together and passed to the Mix Onto element to have either tartrazine or RO water mixed onto the opposite liquids in place to achieve the desired final target concentrations.

Wells within the first 6 columns of each assay plate had tartrazine dispensed into them, with columns 1-3 receiving 100 uL and columns 4-6 receiving 50 uL. Wells on the second half of each assay plate had water dispensed into them, with columns 7-9 receiving 100 uL and columns 10-12 receiving 150 uL. For all transfers multichanneling was enabled, but tip changes were applied when dispensing different volumes and liquids.

Next, final concentrations of 0.05 mM and 0.025 mM tartrazine were achieved by mixing either water or 0.1 M tartrazine onto the liquids in place from the previous steps. 100 uL water was mixed into the wells of columns 1-3 to generate Mix 1, 150 uL water was mixed into wells of columns 4-6 to generate Mix 2, 100 uL 0.1 M tartrazine was mixed into wells of columns 7-9 to generate Mix 3 and 150 uL 0.1 M tartrazine was mixed into wells of columns 10-12 to generate Mix 4. For mixing a custom liquid handling policy was chosen, which performed 5 post mixes using a volume of 100 uL.

In this way 3 replicates per tip/channel were generated for each mixing order and concentration, resulting in 24 replicates for mixing order, and 48 replicates for target concentrations. Since the mixes were performed over two plates there were 96 replicates for each target concentration tested. Additionally, dry versus wet transfers of a sample liquid can be compared, since the sample liquid in this workflow (0.1 M Tartrazine) was dispensed into dry wells for columns 1-6, whereas for columns 7-12 the sample liquid was dispensed onto a liquid already in place (water). This workflow was executed over three devices, namely a Gilson PIPETMAX 268, a Hamilton Microlab STARStar and a Tecan Freedom EVO.

Figure 2: Execution Details Preview in Synthace. Users can simulate, preview and verify their experiments in silico prior to physical execution demonstrated here on the Tecan Freedom EVO.

Labware information and estimated execution run times for each of the liquid handling devices is shown in Table 1. The workflow was run on all three devices on the same day Tartrazine absorbance was measured using a BMG Labtech CLARIOstar microplate reader at 425 nm (Tartrazine 𝜆max), and baseline subtraction performed with absorbance at 620 nm. Data was processed and graphs were generated by a python script in Jupyter Notebook and can be accessed from the Downloads section at the end of this article.

Table 1: Liquid handling information

Independent measurements, and intra- and interplate variability for mixing liquids onto previously aliquot liquids with different concentrations of tartrazine as one of the components of the mixtures were calculated for samples prepared across the different devices tested (Figure 3).

The average intraplate coefficient of variance (CV) was 4.36% (Gilson PIPETMAX 268), 1.92% (Hamilton Microlab STAR), 1.99% (Tecan Freedom EVO). The average interplate CV was 1.22% (Gilson PIPETMAX 268), 0.51% (Hamilton Microlab STAR) and 0.26% (Tecan Freedom EVO).

Figure 3: Absorbance data for mixtures prepared across three devices. Corrected absorbance values of tartrazine for individual mixtures with different target concentrations and mixture orders made by liquid handlers; (A) Gilson PIPETMAX 268, (B) Hamilton Microlab STAR and (C) Tecan Freedom EVO. Each data point represents an individual replicate for the mixture made per plate.This experiment investigated 4 different mix orders: Mix 1 - 100 uL of tartrazine then 100 uL of water mixed onto it. Mix 2 - 50 uL of tartrazine then 150 uL of water mixed onto it. Mix 3 - 100 uL of water then 100 uL of tartrazine mixed onto it. Mix 4 - 150 uL of water then 50 uL of tartrazine mixed onto it..Coefficient of variation (% CV) for each set of replicates on each plate, (D) intraplate variation, and across both plates (E) interplate variation. Dashed horizontal lines represent Synthace internal validation standards for the % CV maximum thresholds.

The following file contains the workflow that was discussed in this technical note. To download the file, right-click the button, then select Save as.

After you download the file, complete the following steps.

The following file contains the raw data that was discussed in this technical note. To download the file, right-click a button, then select Save as.

If you want to process the raw data in the way that was discussed in this technical note, download the Jupyter Notebook or Python scripts. You must edit the scripts to point at the directory that contains the raw data.