Pressure-based microvolume dispensing for high-throughput screening

From LabAutopedia

Jump to: navigation, search
Invited-icon.jpgA LabAutopedia invited article

Authored by: Paul Held Ph.D, BioTek Instruments Inc.

Originally published as a presentation by BioTek Instruments, Inc., September 20, 2006.  Reprinted with permission.



An approach based on the use of positive pressure together with a manifold of precision valving offers the advantage of simplicity, ease of maintenance, and lower cost. This article presents accuracy and precision data for such a dispensing approach, using the NanoQuot™ Microplate Dispenser (Figure 1) from BioTek Instruments (Winooski, VT), which can be employed stand-alone in a pilot assay laboratory or computer controlled as part of an HTS laboratory’s robotics system.
Figure 1. NanoQuot Positive Pressure Nanoliter Dispenser.

Microvolume Dispensing Options

(See also Sample Transport Technologies, Microfluidic Sample Transport)

There are many approaches that can be used for microvolume reagent dispensing into microplates. Contact dispensing, where the dispenser comes in contact with the surface where the liquid is being dispensed, may include syringe technology. Syringe- or aspirate/dispense-based microvolume dispensing is a familiar technology, and when used in microvolumes, air displacement in the tip or probe may be influenced by temperature, pressure, and evaporation to skew the actual volume dispensed. Reusable tips to counter these effects and slow actuation of the syringe piston are often extremely expensive and specific to a particular instrument brand.

Non-contact dispensers eject fluids, thereby eliminating cross-contamination and disruption of the surface where the liquid is being dispensed. Peristaltic pump-based technologies are familiar to many users and are an appealing choice when conservation of reagents is desired or sterility is of acute concern. However in microvolumes, cohesive forces in the liquid are amplified, resulting in uneven distribution of the liquid and notably decreased repeatability. Piezoelectric-based microvolume dispensers use a transducer to create pressure waves in a fluid that form and eject precisely metered droplets. Although highly precise and repeatable, piezoelectric-based microvolume dispensers are often cost-prohibitive for many research applications. Positive pressure dispensers drive solenoid valves to dispense exact and reproducible microvolumes by reacting to fluid viscosity and varying pressure and timing. These positive pressure dispensers tend to be more user-friendly and cost-effective with full control of dispense accuracy.

Pressure-Based Microvolume Dispenser

A positive gas pressure system, in conjunction with eight precision solenoid valves, provides accurate and precise dispensing of volumes ranging from 100 nL to 40 μL in 96-, 384-, and 1536-well plates (Figure 2). The dispense head and tubing set is easily removed and replaced, allowing the nanoliter dispenser to be used for multiple applications without risk of cross-contamination. The Z-axis height adjustment allows for use of half-height plates. As the positive gas pressure system does not require a high input pressure, 30 PSI house air, purified pressured gas, or a dedicated small compressor may be used.

Nanoliter Fig2.jpg 
Figure 2. Fluid path design of a nanoliter dispenser. Positive pressure drives fluids from the reservoir through a distribution manifold, which divides the fluid into 8 separate tubes. Eight precision valves control the fluid dispensed into wells of the microplate. Pressure of the reservoir is monitored by a pressure gauge linked by tubing.

Prior to shipment, the instrument is subject to a 13-point calibration procedure on three different liquid types and pre-loaded with liquid profiles. The end user can easily create new profiles with their specific liquids including viscous solutions such as 40% glycerol and 24% PEG-400 or organic solvents like DMSO and acetonitrile. An uncomplicated user interface provides complete programming capabilities from the keypad. The device has a 12 x 12 inch footprint and 8-inch height.

Materials and Methods

Dispense accuracy and precision were determined using a gravimetric method and the absorbance of dye solutions respectively. Determinations using the gravimetric method were performed by entire plates using an analytical balance. After dispensing fluid to the microplate using the NanoQuot dispenser, the microplate was quickly re-weighed. The resultant weight change, when divided by the number of wells, returns an average per-well dispense volume when corrected for solvent specific gravity. When calculating the precision of dispense using the dye method, an aqueous solution containing FD&C blue #1 dye was dispensed into plates with the nanoliter dispenser. Diluent was then dispensed using a Precision™ XS Microplate Sample Processor (BioTek Instruments, Winooski, VT) and the absorbance at 630 nm (450 reference) was measured using a Synergy™ 2 or Synergy HT Multi-Detection Microplate Reader (BioTek Instruments, Winooski, VT). The resultant absorbance values were then used to calculate dispense precision.


Several different volumes of deionized water were dispensed into 384-well microplates. As demonstrated in Figure 3, the precision of dispensing as measured by %CV was very good across the entire volume range tested. The coefficient of variation (%CV) was always below 5% and for volumes of 0.5 μL or greater the %CV were found to be less than 3% for volumes up to 40 μL. When various volumes of an aqueous dye solution from 100 nL to 2 μL were dispensed and the mean absorbance of each microplate was plotted, a linear relationship is observed (Figure 4). These data indicate that the nanoliter dispenser can reliably dispense across much of its volume range using a single calibration.

Nanoliter Fig3.jpg
Figure 3. Dispense Precision. Various volumes of deionized water containing blue tracer dye were dispensed into 384-well microplates using a NanoQuot nanoliter dispenser. After adding deionized water diluent to a total volume of 100 μL, the absorbance of the wells of the microplate was determined using a Synergy HT Multi-Detection Microplate Reader. The absorbance data was exported to a Microsoft Excel spreadsheet and the mean, standard deviation and %CV calculated. The %CV for each plate was then plotted.
Nanoliter Fig4.jpg
Figure 4. Linearity of Dispense. Different volumes of aqueous dye solutions were dispensed into wells of a 384-well microplate and the absorbance of each well was measured. The data for each plate was then exported to a Microsoft Excel spreadsheet and plotted. Each data point represents the mean of 384 determinations.

Dispensing the lowest allowable volume range is often the most difficult for liquid handlers. As demonstrated in Figure 5, a nanoliter dispenser can consistently dispense 100 nL to all the wells of a 384-well microplate. There is little variation between different valves and little variation between dispenses by the same valve. This is corroborated by the data presented in Figure 6, which depicts a 500 nL dispense of DMSO. These data are presented as the output of each individual valve. The uniformity between valves is indicated by the similarity of all the data lines. Sharp spikes would indicate deviations within any one valve either upward or downward at a particular dispense number.

Nanoliter Fig5.jpg
Figure 5. Uniformity of Dispense. A NanoQuot nanoliter dispenser was used to dispense 100 nL of aqueous dye solution into 384-well microplates. After the addition of dye solution, 100 μL of diluent was added and the absorbance for each well determined. Data for the entire plate was plotted using a Microsoft Excel spreadsheet.
Nanoliter Fig6.jpg
Figure 6. Valve-to-Valve Uniformity. DMSO containing a dye tracer (500 nL) was dispensed into a 384-well plate. After dispensing, 100 μL of water was added and the absorbance at 630 nm of each well was measured using a Synergy HT Multi-Detection Reader. The data was then exported to a Microsoft Excel spreadsheet and each valve tube’s dispense was plotted separately.

The effect of lowering the positive pressure was also examined. There are several different plate types and styles whose geometry can preclude the use of the maximal dispense pressure. The well depth of low profile microplates is approximately half that of a normal 384-well microplate. Use of maximal pressure can lead to splashing after dispensing into the well. In addition, round bottom 384-well microplates are also prone to splashing. Table 1 demonstrates the precision performance of the positive pressure NanoQuot dispenser. Despite using a positive pressure of 10 PSI in lieu of the recommended 15 PSI, the nanoliter dispenser tested provides very acceptable precision for the solutions tested.

Several different solutions containing a tracer dye were dispensed into 384-well microplates with the nanoliter dispenser using positive pressure of 10 PSI. After the addition of water diluent, the absorbance was measured using a Synergy 2 Multi-Detection Microplate Reader and the %CV calculated from the subsequent absorbance values. These data represent the mean and standard deviation calculated from 384 determinations.

Table 1. Dispense Precision for Various Solutions at 10 PSI Positive Pressure
  Dispense Volume (μL)
Solution 0.5 1.0 2.0
100% Ethanol 4.2* 6.7 5.8
70% Ethanol 5.1 4.3 4.1
50% Ethanol 7.4 7.1 3.7
PBS 6.7 5.6 3.8
Acetonitrile 4.8 4.1 5.1
10% DMSO 4.8 4.3 3.4
98% DMSO 8.7 6.9 5.0
1% BSA 8.9 7.8 3.7
10% Glycerol 8.1 8.7 3.9
*Data represent the %CV across a 384-well plate 

Under some circumstances, multiple dispenses of reagents, such as multiple reagents or repetitive dispenses of the same reagent, are made into the same well. These types of dispenses are prone to splashing, due to the shortened distance from the dispense valve tip and the fluid surface, as well as the potential for the existing fluid to be projected out of the well. As demonstrated in Table 2, despite reducing the pressure to 7.5 PSI, the accuracy of the nanoliter dispenser after a total of three dispenses (5 μL, 5 μL and 10 μL) of a sodium fluorescein solution into a round bottom 384-well microplate was quite good. The combined error of the 3 dispenses was less than 3% and the overall precision resulted in a %CV of approximately 5%. The low positive pressure was required to minimize splashing resulting from both the rounded surface of the well bottom and the partially filled nature of the well itself.

Table 2. Accuracy and Precision with Multiple Dispenses
  Accuracy Precision
Expected Total Vol Actual Vol (μL)  % Error Mean  %CV
20 μL 19.48* -2.60% 95854** 5.8
20 μL 19.57 -2.15% 97775 5.2
*Accuracy measurements represent the calculated per-well volume as determined gravimetrically after a total of 3 dispenses of 5, 5, and 10 μL respectively.
**Precision measurements represent the mean of 384-well fluorescent determinations of sodium fluorescein.

September 20, 2006

Paul Held Ph.D., Staff Scientist, Applications Department
Wendy Goodrich, Applications Scientist, Applications Department
Jason Greene, Product Manager
BioTek Instruments, Inc.
Winooski, VT, USA

Click [+] for other articles on 
Liquid handling(3 C, 12 P)
The Market Place for Lab Automation & Screening  Automated Liquid Handling Systems Liquid Handling Components OEM
Click [+] for other articles on 
The Market Place for Lab Automation & Screening  Microplate Dispensers
Click [+] for other articles on 
High Throughput Screening(1 C, 15 P)
The Market Place for Lab Automation & Screening  Drug Discovery HTS Products