Application Note 25 Making optical density measurements automatically corrected to a 1 cm pathlength in the SPECTRAmax PLUS microplate spectrophotometer (The three P s of Pathcheck : Principle, Procedures and Precautions) TABLE OF CONTENTS INTRODUCTION.................................................. 2 PRINCIPLES OF PATHCHECK...................................... 2 Pathlength correction using near infra-red measurements............. 2 The pre-installed Water Constant: a shortcut........................ 3 PATHLENGTH CORRECTION BY THE SPECTRAmax PLUS............ 3 Two ways to correct pathlength................................... 3 Why must the plate be pre-read?.................................. 4 USING PATHCHECK............................................... 4 Getting ready................................................... 4 Materials..................................................... 4 Set up SOFTmax PRO to use PathCheck............................ 5 Pre-read the microplate.......................................... 6 Pathlength correction using a cuvette reference...................... 7 Pathlength correction using the Water Constant..................... 7 Using PathCheck to inspect a microplate for pipetting errors.......... 8 EXPERIMENTAL RESULTS.......................................... 9 Experimental design and cuvette results............................ 9 Plate optical densities displayed without pathlength correction....... 11 Plate optical densities displayed with pathlength correction.......... 12 Plate data displayed as absolute pathlength in centimeters........... 13 The three data presentations combined into a single graph........... 13 DISCUSSION..................................................... 14 Advice and precautions......................................... 14 Evaporation.................................................. 14 Failure to subtract appropriate pre-read values...................... 14 Potential interference........................................... 15 1
INTRODUCTION UV/VIS spectrophotometers and microplate readers differ fundamentally in their beam geometry. In spectrophotometers, samples are read through cuvettes or tubes with a horizontal (cross-sectional) light path. The horizontal light beam and the customary 1 cm pathlength make assays based on extinction coefficients straightforward and allow easy comparison of results between labs. In microplate readers, the vertical light beam results in a pathlength that depends on the volume of fluid in each well. The variable pathlength in microplates has made it difficult to perform extinctionbased assays and confusing to compare results obtained in a microplate reader with those obtained in a spectrophotometer. This problem has been remedied by the introduction of the SPECTRAmax PLUS microplate spectrophotometer and its PathCheck feature, the ability to determine the pathlength in each well of a microplate and automatically normalize the absorbance value to a 1 cm pathlength. This application note discusses the principles on which PathCheck is based, and gives specific instructions for using it with a SPECTRAmax PLUS microplate spectrophotometer and SOFTmax PRO 2.0 software. PRINCIPLES OF PATHCHECK Pathlength correction using near infra-red measurements Water is essentially transparent from 200 nm to 900 nm, but in the near infrared (NIR), it has a distinctive absorption peak near 977 nm (Figure 1). This characteristic absorbance band in a spectral region where most biological molecules have little or no absorbance can be utilized to measure the pathlength of light through an aqueous sample. 0.26 0.24 0.22 0.20 Absorbance 0.18 0.16 0.14 0.12 0.10 0.08 0.06 0.04 0.02 900 920 940 960 980 1000 1020 1040 1060 Wavelength (nm) Figure 1: Absorbance of water Lambert s law of light absorption predicts absorbance is proportional to the distance that light travels through the sample: the longer the pathlength, the higher the absorbance peak. Baseline absorbance can be measured at a wavelength distant from the water peak, e.g. 900 nm. The pathlength of light 2
through an aqueous sample can be calculated by comparing the peak height obtained in a microplate well with the peak height obtained in a standard 1 cm cuvette: ( A 977 A 900 ) sample ( -------------------------------------------------------------------------------------------- = A 977 A 900 ) Pathlength of aqueous sample 1.0 cm aqueous solvent In practice, 977 nm is not an ideal wavelength for pathlength measurements because at that wavelength the absorption of water is temperature-dependent. Pathlength measurements are subject to error if the microplate and cuvette measurements are not made at exactly the same temperature. SPECTRAmax PLUS avoids temperature-dependency by measuring water absorbance at 1000 nm, near a temperature isosbestic point. The trade-off is an increased requirement for the instrument to have superior wavelength reproducibility, because the measurement is made on a slope of the absorbance curve, rather than at the absorbance maximum. Using the ratio in the equation above, the absorbance of the aqueous sample in a microplate well (in the example below, protein at 280 nm) can be corrected to a 1 cm pathlength as follows: ( A 280 ) ( A sample 1000 A 900 ) ------------------------------------------------------------------------------------------------------------------------------------------- 1.0 cm aqueous solvent = ( A 1000 A 900 ) A 280 sample corrected to 1.0 cm sample The pre-installed Water Constant: a shortcut Because comparatively few substances absorb in the NIR, the quantity (A 1000 A 900 ) 1.0 cm aqueous solvent in the equation above is, for aqueous solutions, practically a constant, and is referred to as the Water Constant. The value of the Water Constant for each SPECTRAmax PLUS is determined during instrument manufacture and is pre-installed into the instrument s firmware. Substituting the Water Constant for (A 1000 A 900 ) 1.0 cm aqueous solvent in the equation for correcting absorbance values in the UV/VIS region then yields: ( A 280 ) Water Constant sample -------------------------------------------------------------------------------- = ( A 1000 A 900 ) A 280 sample corrected to 1.0 cm sample PATHLENGTH CORRECTION BY THE SPECTRAMAX PLUS Two ways to correct pathlength The SPECTRAmax PLUS gives you two options when making pathlength-corrected measurements: 1) putting the aqueous solvent into a clean 1 cm reference cuvette in the instrument s cuvette port and having the instrument base the calculations on the A 1000 and A 900 of the cuvette, or 2) using the Water Constant. the first option is strongly recommended because it ensures the highest possible accuracy and will accommodate a variety of sample solvents. Indeed, a cuvette reference is necessary in cases where the sample solvent is greater than 0.5 M in salts or solutes, contains organic solvent or has unusual absorbance in the NIR. Use of the Water Constant should be considered a shortcut and potentially less accurate. 3
In addition to reporting absorbance measurements corrected to a 1 cm pathlength, SOFTmax PRO can report the pathlength in each of the 96 wells. This ability is useful to screen a microplate for volume irregularities or to check for pipetting errors. When performing an absorbance read with PathCheck, SOFTmax PRO performs the following sequence of calculations: 1 Subtraction of the plate pre-read values from the normal read values (i.e., subtraction of the plate background in the UV/VIS region). 2 Pathlength correction (using the ratio of the NIR measurements). 3 Subtraction of plate/group blanks (if any). 4 Additional user-specified custom data reduction (if any). Why must the plate be pre-read? The plate pre-read doesn t necessarily correct for well-to-well variability of the microplate. Instead, the purpose is to subtract the absorbance due to the plate material before pathlength correction is applied. The NIR measurements for pathlength correction do not need blanking or pre-reading because the calculation requires the difference between A 1000 and A 900, and the plate background is essentially the same at both wavelengths. However, at the wavelength(s) in the UV/VIS where the analyte is measured, the plate can contribute significantly to the overall absorbance, therefore the absorbance due to the plate must be subtracted before pathlength correction is applied. For accurate pathlength-corrected measurements, the plate must be pre-read with water in the wells. The reason is not because water absorbs (it doesn t), but because the refractive index of light differs at a plate/air interface versus a plate/water interface. If the plates are sufficiently transparent at the wavelength(s) of interest and have low well-to-well variability, the pre-read doesn t have to be made on the same plate the samples will be read in; another clean plate from the same lot will be adequate. Even more convenient, a single plate can be pre-read, then the pre-read data stored for subsequent use. On the other hand, if the plates are inherently variable at the wavelength(s) of interest, it may be necessary to use the same plate for pre-reading and reading samples, so that adequate well-by-well correction is achieved. USING PATHCHECK Getting ready Materials 1 SPECTRAmax PLUS microplate spectrophotometer 2 High-quality microplates; e.g.: SPECTRAplate Quartz, suitable for all wavelengths, Molecular Devices Corp. (catalog # R8024) SPECTRAplate Quartz half-area microplate, suitable for all wavelengths, Molecular Devices Corp. (catalog # R8028) Quartz microtest 8-well strip, suitable for all wavelengths, Hellma (catalog # 730.010QS); use with a frame for 8-well strips, E & K Scientific Products (catalog # 564101) Note: the 8-well strip frame sold by Evergreen Scientific does not fit 4
properly. Disposable UV Plate, suitable for wavelengths > 220 nm (pre-read on same plate recommended below 250 nm), Corning Costar Corp. (catalog # 3635) Nunc flatbottom plates, suitable for wavelengths > 320 nm (pre-read on same plate recommended below 350 nm), Fisher Scientific (catalog # 12565226) Greiner flatbottom plates, suitable for wavelengths > 400 nm (340 nm to 400 nm with same-plate pre-read), E & K Scientific Products (catalog # 565101) 3 Pipettor and tips or transfer pipets suitable for use with microplates 4 Samples (100-300 µl each) Set up SOFTmax PRO to use PathCheck Step 1 Step 2 Launch SOFTmax PRO, then open a Plate Section or, if necessary, create a new Plate Section. Set up the Instrument Settings dialog box as shown in Figure 2. Select to perform an endpoint read at the desired wavelength(s). Click the Path- Check box to mark it, (pre-read plate will automatically be checked) then click the Cuvette Reference or Water Constant button, depending upon which you wish to use. Automix is not advised because it provides no benefit in this measurement. If you wish to read only part of the plate, click the Strips button and select the strips (columns) to be read. Figure 2: The Instrument Settings dialog box, set up for measurements incorporating PathCheck Step 3 Step 4 Use the Template Editor to create a template showing where standards and unknowns will be located on the microplate. Click the Reduction button in the toolbar to display the Reduction dialog box (Figure 3). Set the wavelength combination (for example, to L1 for single wavelength reads) and the data mode to Absorbance. Click in 5
the Apply Pathcheck checkbox to select it, upon which SOFTmax PRO will automatically select Use pre-read plate. Figure 3: The Reduction dialog box, set up for applying a pre-read and PathCheck Pre-read the microplate Step 1 Inspect a clean microplate by holding it up to the light and looking for dust. If needed, blow clean air into and across it, to remove dust particles. Note: For highest accuracy it is important that the plate be dust-free, because dust particles can interfere with the pre-read. Step 2 Step 3 Step 4 Step 5 Pipet distilled water into each well. (The volumes should be about the same as the subsequent sample volumes though accurate pipetting is not necessary.) Put the plate in the SPECTRAmax PLUS drawer, then click the Read button in the Plate section s Status Bar. A dialog box will appear, requesting you to confirm that the read is a pre-read. Click the OK button. After the pre-read is complete, remove the plate from the drawer, pour out the water and blot it dry. If you are planning to use the pre-read data for more than one plate, save an extra copy of the data file containing the pre-read values. Quartz plates are so uniform that you do not need to subtract the plate background on a well-by-well basis. You can use different plates for the pre-reads and sample reads. Better still, you can pre-read once and store the pre-read file for subsequent re-use, being careful to use the Save As command to save the file with a different name each time you use it, to avoid over-writing the original file. Nunc flat-bottom plates are also sufficiently uniform that stored preread values can be used above 380 nm. Between 320 and 380 nm, it is advisable to use the same plate for pre-reads and sample reads. If using Costar UV plates, you may or may not choose to use stored preread values obtained from a plate from the same lot, depending on the precision you require. Costar plates have approximately.050 OD with approximately.007 OD spread at 280 nm. If the raw absorbance in a given well is 1.0, the potential error due plate background variation is less than 1%. But if the raw absorbance is 0.1, the maximum potential error in a given well approaches 10%. Between 220 nm and 250 nm, it is advisable to use the same plate for pre-reads and sample reads. 6
Pathlength correction using a cuvette reference If you are using stored pre-read values (see above), open the data file containing the pre-read data you want to use, then skip Steps 1 & 2 below. Step 1 Step 2 Step 3 Step 4 Step 5 Set up SOFTmax PRO to use PathCheck as detailed in "Set up SOFTmax PRO to use PathCheck" on page 5. Select Read Cuvette Reference in the Instrument Settings dialog box. Pre-read an appropriate microplate, following the instructions above. Place a clean quartz or glass cuvette containing distilled water or your aqueous sample solvent in the SPECTRAmax PLUS cuvette port. Plastic cuvettes are also generally acceptable, but the performance of a particular cuvette should be verified before use in an assay. Take the pre-read microplate (or another microplate from the same high quality lot) and transfer aliquots of your samples and blanks into their designated wells. Accurate pipetting is not necessary though the volumes should be between 100 µl and 300 µl for best pathlength-corrected results. Put the plate in the SPECTRAmax PLUS drawer, then click the Read button in the Plate section s Status Bar. A dialog box will appear, requesting you to confirm that the read is a normal read. Click the OK button. The cuvette will be read at the same time the plate is read, and its NIR absorbance values used in the PathCheck calculations. Note:To avoid errors due to evaporation, make the readings within a few minutes of putting the samples in the microplate. If the read must be delayed, cover the plate with an adhesive seal. Remove the seal immediately prior to reading the plate. Pathlength correction using the Water Constant Step 1 Set up SOFTmax PRO to use PathCheck as detailed in "Set up SOFTmax PRO to use PathCheck" on page 5. Select Water Constant in the Instrument Settings dialog box (Figure 4). Figure 4: The Instrument Settings dialog box, set up to use PathCheck with the Water Constant Step 2 Pre-read an appropriate microplate, following the instructions above. 7
Step 3 Step 4 Take the pre-read microplate (or another microplate from the same high quality lot) and transfer aliquots of your samples and blanks into their designated wells. Accurate pipetting is not necessary, though the volumes should be between 100 µl and 300 µl for best pathlength-corrected results. Put the plate in the SPECTRAmax PLUS drawer. Click the Read button in the Plate section s Status Bar. A dialog box will appear, requesting you to confirm that the read is a normal read. Click the OK button. Note:To avoid errors due to evaporation, make the readings within a few minutes of putting the samples in the microplate. If the read must be delayed, cover the plate with an adhesive seal. Remove the seal immediately prior to reading the plate. Using PathCheck to inspect a microplate for pipetting errors Step 1 Step 2 Using SOFTmax PRO, select a Plate Section or, if necessary, create a new Plate by selecting New Plate from the Experiment menu. Set up the Instrument Settings dialog box as shown in Figure 5. Select to perform an end-point read at any wavelength. Note: SOFTmax PRO requires that you enter a UV/VIS wavelength at which to read, even though that measurement is not used in calculating pathlength Select PathCheck with a cuvette reference (or use the Water Constant, if desired) and de-select pre-read plate. (Pre-reading is unnecessary because the UV/VIS measurement is not used in calculating pathlength.) Figure 5: The Instrument Settings dialog box, set up to check pipetting accuracy with PathCheck Step 3 Click the Reduction button in the Plate Section s tool bar to display the Reduction dialog box (Figure 6). Select Custom from the Wavelength Combination pop-up menu, then click the formula button to display the Calculation dialog box. Type!Pathlength in the formula field of the 8
dialog box (do not include the quote marks, and do not put a space between the! and Pathlength). Figure 6: The Reduction dialog box, set up for checking pipetting accuracy with PathCheck Step 4 For quick visualization of the plate it is useful to display the data in grayscale. Click the Display button in the Plate section s toolbar to produce the Display dialog box (Figure 7), then click the Grayscale button. If desired, you can also select to have the reduced number displayed in each well. Figure 7: The Display dialog box, set to display data in grayscale Step 5 Put the plate into the of SPECTRAmax PLUS drawer, then click the Read button in the Plate Status Bar. EXPERIMENTAL RESULTS Experimental design and cuvette results A yellow buffer solution (Fisher certified buffer, ph 7.00) was used to illustrate the performance of SPECTRAmax PLUS and PathCheck. The buffer has an absorbance maximum at 426 nm and its absorbance at that wavelength was first measured in a standard 1 cm cuvette. The buffer was then pipetted into wells of a Nunc flat-bottomed microplate in volumes of 75 to 300 µl, and the absorbance was measured, using PathCheck to normalize it to a 1 cm pathlength. Note: 75 µl is less than the recommended minimum volume per well, but was included to illustrate the variability occurring at marginal volumes. 9
A Cuvette Set was created and in the Cuvette Set Template dialog box, a group called Cuvette was created, comprising 3 samples named Yellow Buffer (Figure 8). Figure 8: The Template Editor for a Cuvette Section In the Instrument Setup dialog box for the Cuvette Set, an endpoint read at 426 nm was specified. A glass cuvette was filled with deionized water and placed in the SPECTRAmax PLUS cuvette port. The REF button in the status bar was clicked to perform a reference read on the cuvette. The cuvette was then replaced with a matched cuvette containing the yellow buffer, and the absorbance of the yellow buffer was read three times. Figure 9 shows the resulting absorbance values in the Cuvette Set and in the Group Table. The results indicate that the absorbance of the yellow buffer at 426 nm in a standard 1 cm cuvette was 0.386. Figure 9: The Cuvette set for Fisher yellow buffer, and the corresponding Group Table 10
A Nunc flat-bottom microplate was pre-read, then the buffer was pipetted into a second microplate from the same lot, in the volumes specified in the template (Figure 10). The cuvette was left in the cuvette port and the plate was read using the cuvette as a reference. Figure 10: Template showing the location of the sample volumes on the microplate Plate optical densities displayed without pathlength correction Figure 11 displays the results when PathCheck is de-selected in the Reduction dialog box. The plate section is displayed in grayscale for easy visualization. As one would predict, without pathlength correction, the absorbance values increase with increasing volume of sample in the microplate wells. The Group Table indicates that the mean values range from 0.09 for the 75 µl group to 0.336 for the 300 µl group. Figure 11: Plate Section and Group Table showing plate data with PathCheck deselected 11
Plate optical densities displayed with pathlength correction Figure 12 shows the same set of data with PathCheck applied to it. For all the sample volumes, the corrected values are close to the cuvette value. The column Wells/Cuvette in the Group Table gives the mean corrected OD 426 divided by the value obtained in the cuvette (0.386). The mean values are all within 1% of the cuvette value. The considerably greater variability in the wells containing only 75 µl is obvious in the grayscale representation. Even though the latter group was still within 1% of the cuvette value in this carefully controlled experiment, we recommend that PathCheck be applied only to wells containing at least 100 µl. Figure 12: Plate Section and Group Table showing plate data with PathCheck applied 12
Plate data displayed as absolute pathlength in centimeters Figure 13 shows the same data set displayed in absolute pathlength. Not surprisingly, the grayscale representation of the data looks similar to that in Figure 11, which displays the raw OD 426 values. The Group Table indicates that the mean pathlength values ranged from 0.231 cm to 0.874 cm. Figure 13: Plate Section and Group Table showing mean pathlength in centimeters The three data presentations combined into a single graph In Figure 12, absorbance at 426 nm with and without pathlength correction is plotted as a function of pathlength in centimeters. As expected, the uncorrected absorbance values increase with increasing pathlength (i.e., with increasing volume in the wells). When PathCheck is applied, the corrected absorbance values are almost identical. Figure 14: OD 426 with and without pathlength correction 13
DISCUSSION Advice and precautions When pathlength-corrected measurements are made using an appropriate procedure, the values should be within 2.5% or better of values obtained from the same solution in a 1 cm cuvette. Two potential causes of erroneous results are evaporation and failure to subtract pre-read values. Evaporation Recall the equation for pathlength correction: ( A 280 ) ( A sample 1000 A 900 ) ------------------------------------------------------------------------------------------------------------------------------------------- 1.0 cm aqueous solvent = ( A 1000 A 900 ) A 280 sample corrected to 1.0 cm sample If evaporation occurs, the value for (A 1000 - A 900 ) sample in the denominator decreases, causing a progressive overcorrection. (The reading in the UV/VIS region does not change unless the analyte happens to be volatile.) The effect is particularly noticeable in wells with small volumes. Even if the wells contain 300 µl of solution, the effect of evaporation is noticeable within 15 min (~1% error). To avoid errors due to evaporation, take the readings within minutes of putting the samples in the microplate. If the read must be delayed, cover the plate with an adhesive seal, then remove the seal immediately before reading the plate. Failure to subtract appropriate pre-read values When using PathCheck to correct for pathlength through a sample, SOFTmax PRO goes through the following sequence of calculations: 1 Subtraction of the plate pre-read values from the normal read values (i.e., the readings in UV/VIS region). 2 Pathlength correction (using the ratio of the NIR measurements). 3 Subtraction of plate/group blanks (if any). 4 Additional user-specified custom data reduction (if any). The purpose of the pre-read is not necessarily to correct for well-to-well variability in the microplate, but instead to subtract the absorbance due to the plate material before pathlength correction is applied. The plate should be preread with water in the wells. Wells pre-read with water generally have less absorbance than dry wells. (e.g., at 426 nm Nunc microplates give absorbance values of approximately 0.037 and 0.049 when read with and without water in the wells.) Thus pre-reading a dry plate results in pathlength-corrected absorbance values that are lower than the cuvette values. If the pre-read step is omitted entirely, the pathlength-corrected results are grossly high. The good news is that pre-reads do not necessarily have to be made with the same plate as the final reads. Some plates are so uniform (2-3 MOD spread overall) that one plate can be pre-read and another clean, dry plate be used for the samples. Even more convenient, pre-read data can be stored in SOFTmax PRO and used over and over again. Also, SOFTmax PRO allows you to copy and paste pre-read data from one plate section to another, provided the instrument settings are identical in the two plate sections. 14
Potential interference Pathlength correction is intended for use with aqueous solutions, although small quantities of organic solvents can be tolerated if the PathCheck measurements are made with a cuvette reference. As stated above, a cuvette reference is strongly recommended for pathlength-corrected measurements, although the water constant can often be used as a shortcut. However, if the sample solvent contains a high concentration (e.g. > 0.5 M) of salts or other solutes, or contains organic solvent, the NIR absorbance values may differ from water. In such cases the Water Constant is not appropriate, and it is essential to take pathlength-corrected measurements using a cuvette reference containing the sample solvent. Pathlength correction using NIR absorbance is appropriate as long as there is nothing in the sample that interferes with the measurements. Turbidity is one source of interference. Also, any molecule with differential absorption at 1000 and 900 nm will interfere if present in a high enough concentration. Although few biological substances absorb between 900 and 1000 nm, interference may be expected from some highly-conjugated molecules such as the pthalocyanines, chlorophylls, carotenoids, phycobilins, as well as porphyrin-containing or related molecules such a myoglobin, hemoglobin and peroxidases. Reduced phosphomolybdate complexes also absorb between 900 and 1000 nm. Pathlength correction in such solutions, however, can still be applied if the interfering substances are dilute enough such their NIR absorption is insignificant relative to water. Provided that the interfering species is present at a constant concentration in all samples, the interference also can be eliminated by placing a sample with the interfering species in the reference cuvette. SALES OFFICE United States & Canada Molecular Devices Tel. +1-800-635-5577 Fax +1-408-747-3601 Brazil Molecular Devices Brazil Tel. +55-11-3616-6607 Fax +55-11-3616-6607 China Molecular Devices Shanghai Tel. +86-21-6887-8820 Fax +86-21-6887-8890 Germany Molecular Devices GmbH Tel. +49-89/96-05-88-0 Fax +49-89/9-62-02-34-5 Japan Molecular Devices Japan, Osaka Tel. +81-6-6399-8211 Fax +81-6-6399-8212 Molecular Devices Japan, Tokyo Tel. +81-3-5282-5261 Fax +81-3-5282-5262 South Korea Molecular Devices Korea, LLC Tel. +82-2-3471-9531 Fax +82-2-3471-9532 United Kingdom Molecular Devices Ltd. Tel. +44-118-944-8000 Fax +44-118-944-8001 www.moleculardevices.com FOR RESEARCH USE ONLY. NOT FOR USE IN DIAGNOSTIC PROCEDURES. The trademarks used herein are the property of Molecular Devices, Inc. or their respective owners. 2010 Molecular Devices, Inc. Printed in U.S.A. 6/10 #0120-1120B 15