A revolutionary new inline static mixer has been developed specifically designed to meet the stringent requirements of high performance liquid chromatography (HPLC) and ultra high performance liquid chromatography (HPLC and UHPLC) systems. Poor mixing of two or more mobile phases can result in a higher signal-to-noise ratio, which reduces sensitivity. Homogeneous static mixing of two or more fluids with a minimum internal volume and physical dimensions of a static mixer represents the highest standard of an ideal static mixer. The new static mixer achieves this by using new 3D printing technology to create a unique 3D structure that provides improved hydrodynamic static mixing with the highest percentage reduction in base sine wave per unit internal volume of the mixture. Using 1/3 of the internal volume of a conventional mixer reduces the basic sine wave by 98%. The mixer consists of interconnected 3D flow channels with varying cross-sectional areas and path lengths as the fluid traverses complex 3D geometries. Mixing along multiple tortuous flow paths, combined with local turbulence and eddies, results in mixing at the micro, meso and macro scales. This unique mixer is designed using computational fluid dynamics (CFD) simulations. The test data presented show that excellent mixing is achieved with a minimum internal volume.
For more than 30 years, liquid chromatography has been used in many industries, including pharmaceuticals, pesticides, environmental protection, forensics, and chemical analysis. The ability to measure to parts per million or less is critical to technological development in any industry. Poor mixing efficiency leads to poor signal-to-noise ratio, which is an annoyance to the chromatography community in terms of detection limits and sensitivity. When mixing two HPLC solvents, it is sometimes necessary to force mixing by external means to homogenize the two solvents because some solvents do not mix well. If solvents are not thoroughly mixed, degradation of the HPLC chromatogram may occur, manifesting itself as excessive baseline noise and/or poor peak shape. With poor mixing, baseline noise will appear as a sine wave (rising and falling) of the detector signal over time. At the same time, poor mixing can lead to broadening and asymmetric peaks, reducing analytical performance, peak shape, and peak resolution. The industry has recognized that in-line and tee static mixers are a means of improving these limits and allowing users to achieve lower detection limits (sensitivities). The ideal static mixer combines the benefits of high mixing efficiency, low dead volume and low pressure drop with minimum volume and maximum system throughput. In addition, as analysis becomes more complex, analysts must routinely use more polar and difficult-to-mix solvents. This means better mixing is a must for future testing, further increasing the need for superior mixer design and performance.
Mott has recently developed a new range of patented PerfectPeakTM inline static mixers with three internal volumes: 30 µl, 60 µl and 90 µl. These sizes cover the range of volumes and mixing characteristics needed for most HPLC tests where improved mixing and low dispersion are required. All three models are 0.5″ in diameter and deliver industry-leading performance in a compact design. They are made of 316L stainless steel, passivated for inertness, but titanium and other corrosion resistant and chemically inert metal alloys are also available. These mixers have a maximum operating pressure of up to 20,000 psi. On fig. 1a is a photograph of a 60 µl Mott static mixer designed to provide maximum mixing efficiency while using a smaller internal volume than standard mixers of this type. This new static mixer design uses new additive manufacturing technology to create a unique 3D structure that uses less internal flow than any mixer currently used in the chromatography industry to achieve static mixing. Such mixers consist of interconnected three-dimensional flow channels with different cross-sectional areas and different path lengths as the liquid crosses complex geometric barriers inside. On fig. Figure 1b shows a schematic diagram of the new mixer, which uses industry standard 10-32 threaded HPLC compression fittings for inlet and outlet, and has shaded blue borders of the patented internal mixer port. Different cross-sectional areas of the internal flow paths and changes in flow direction within the internal flow volume create regions of turbulent and laminar flow, causing mixing at the micro, meso and macro scales. The design of this unique mixer used computational fluid dynamics (CFD) simulations to analyze flow patterns and refine the design before prototyping for in-house analytical testing and customer field evaluation. Additive manufacturing is the process of printing 3D geometric components directly from CAD drawings without the need for traditional machining (milling machines, lathes, etc.). These new static mixers are designed to be manufactured using this process, where the mixer body is created from CAD drawings and the parts are fabricated (printed) layer by layer using additive manufacturing. Here, a layer of metal powder about 20 microns thick is deposited, and a computer-controlled laser selectively melts and fuses the powder into a solid form. Apply another layer on top of this layer and apply laser sintering. Repeat this process until the part is completely finished. The powder is then removed from the non-laser bonded part, leaving a 3D printed part that matches the original CAD drawing. The final product is somewhat similar to the microfluidic process, with the main difference being that the microfluidic components are usually two-dimensional (flat), while using additive manufacturing, complex flow patterns can be created in three-dimensional geometry. These faucets are currently available as 3D printed parts in 316L stainless steel and titanium. Most metal alloys, polymers and some ceramics can be used to make components using this method and will be considered in future designs/products.
Rice. 1. Photograph (a) and diagram (b) of a 90 μl Mott static mixer showing a cross-section of the mixer fluid flow path shaded in blue.
Run computational fluid dynamics (CFD) simulations of static mixer performance during the design phase to help develop efficient designs and reduce time-consuming and costly trial-and-error experiments. CFD simulation of static mixers and standard piping (no-mixer simulation) using the COMSOL Multiphysics software package. Modeling using pressure-driven laminar fluid mechanics to understand fluid velocity and pressure within a part. This fluid dynamics, combined with the chemical transport of mobile phase compounds, helps to understand the mixing of two different concentrated liquids. The model is studied as a function of time, equal to 10 seconds, for ease of calculation while searching for comparable solutions. Theoretical data were obtained in a time-correlated study using the point probe projection tool, where a point in the middle of the exit was chosen for data collection. The CFD model and experimental tests used two different solvents through a proportional sampling valve and pumping system, resulting in a replacement plug for each solvent in the sampling line. These solvents are then mixed in a static mixer. Figures 2 and 3 show flow simulations through a standard pipe (no mixer) and through a Mott static mixer, respectively. The simulation was run on a straight tube 5 cm long and 0.25 mm ID to demonstrate the concept of alternating plugs of water and pure acetonitrile into the tube in the absence of a static mixer, as shown in Figure 2. The simulation used the exact dimensions of the tube and mixer and a flow rate of 0 .3 ml/min.
Rice. 2. Simulation of CFD flow in a 5 cm tube with an internal diameter of 0.25 mm to represent what happens in an HPLC tube, i.e. in the absence of a mixer. Full red represents the mass fraction of water. Blue represents the lack of water, i.e. pure acetonitrile. Diffusion regions can be seen between alternating plugs of two different liquids.
Rice. 3. Static mixer with a volume of 30 ml, modeled in the COMSOL CFD software package. The legend represents the mass fraction of water in the mixer. Pure water is shown in red and pure acetonitrile in blue. The change in the mass fraction of the simulated water is represented by a change in the color of the mixing of two liquids.
On fig. 4 shows a validation study of the correlation model between mixing efficiency and mixing volume. As the mixing volume increases, the mixing efficiency will increase. To the authors’ knowledge, other complex physical forces acting inside the mixer cannot be accounted for in this CFD model, resulting in higher mixing efficiency in experimental tests. The experimental mixing efficiency was measured as the percentage reduction in the base sinusoid. In addition, increased back pressure usually results in higher mixing levels, which are not taken into account in the simulation.
The following HPLC conditions and test setup were used to measure raw sine waves to compare the relative performance of different static mixers. The diagram in Figure 5 shows a typical HPLC/UHPLC system layout. The static mixer was tested by placing the mixer directly after the pump and before the injector and separation column. Most background sinusoidal measurements are made bypassing the injector and capillary column between the static mixer and the UV detector. When evaluating the signal-to-noise ratio and/or analyzing the peak shape, the system configuration is shown in Figure 5.
Figure 4. Plot of mixing efficiency versus mixing volume for a range of static mixers. The theoretical impurity follows the same trend as the experimental impurity data confirming the validity of the CFD simulations.
The HPLC system used for this test was an Agilent 1100 Series HPLC with a UV detector controlled by a PC running Chemstation software. Table 1 shows typical tuning conditions for measuring mixer efficiency by monitoring basic sinusoids in two case studies. Experimental tests were carried out on two different examples of solvents. The two solvents mixed in case 1 were solvent A (20 mM ammonium acetate in deionized water) and solvent B (80% acetonitrile (ACN)/20% deionized water). In Case 2, solvent A was a solution of 0.05% acetone (label) in deionized water. Solvent B is a mixture of 80/20% methanol and water. In case 1, the pump was set to a flow rate of 0.25 ml/min to 1.0 ml/min, and in case 2, the pump was set to a constant flow rate of 1 ml/min. In both cases, the ratio of the mixture of solvents A and B was 20% A/80% B. The detector was set to 220 nm in case 1, and the maximum absorption of acetone in case 2 was set to a wavelength of 265 nm.
Table 1. HPLC Configurations for Cases 1 and 2 Case 1 Case 2 Pump Speed 0.25 ml/min to 1.0 ml/min 1.0 ml/min Solvent A 20 mM ammonium acetate in deionized water 0.05% Acetone in deionized water Solvent B 80% Acetonitrile (ACN) / 20% deionized water 80% methanol / 20% deionized water Solvent ratio 20% A / 80% B 20% A / 80% B Detector 220 nm 265 nm
Rice. 6. Plots of mixed sine waves measured before and after applying a low-pass filter to remove baseline drift components of the signal.
Figure 6 is a typical example of mixed baseline noise in Case 1, shown as a repeating sinusoidal pattern superimposed on baseline drift. Baseline drift is a slow increase or decrease in the background signal. If the system is not allowed to equilibrate long enough, it will usually fall, but will drift erratically even when the system is completely stable. This baseline drift tends to increase when the system is operating in steep gradient or high back pressure conditions. When this baseline drift is present, it can be difficult to compare results from sample to sample, which can be overcome by applying a low-pass filter to the raw data to filter out these low-frequency variations, thereby providing an oscillation plot with a flat baseline. On fig. Figure 6 also shows a plot of the mixer’s baseline noise after applying a low-pass filter.
After completing the CFD simulations and initial experimental testing, three separate static mixers were subsequently developed using the internal components described above with three internal volumes: 30 µl, 60 µl and 90 µl. This range covers the range of volumes and mixing performance required for low analyte HPLC applications where improved mixing and low dispersion are required to produce low amplitude baselines. On fig. 7 shows basic sine wave measurements obtained on the test system of Example 1 (acetonitrile and ammonium acetate as tracers) with three volumes of static mixers and no mixers installed. Experimental test conditions for the results shown in Figure 7 were held constant throughout all 4 tests according to the procedure outlined in Table 1 at a solvent flow rate of 0.5 ml/min. Apply an offset value to the datasets so they can be displayed side by side without signal overlap. Offset does not affect the amplitude of the signal used to judge the performance level of the mixer. The average sinusoidal amplitude without the mixer was 0.221 mAi, while the amplitudes of the static Mott mixers at 30 µl, 60 µl, and 90 µl dropped to 0.077, 0.017, and 0.004 mAi, respectively.
Figure 7. HPLC UV Detector Signal Offset vs. Time for Case 1 (acetonitrile with ammonium acetate indicator) showing solvent mixing without mixer, 30 µl, 60 µl and 90 µl Mott mixers showing improved mixing (lower signal amplitude ) as the volume of the static mixer increases. (actual data offsets: 0.13 (no mixer), 0.32, 0.4, 0.45mA for better display).
The data shown in fig. 8 are the same as in Fig. 7, but this time they include the results of three commonly used HPLC static mixers with internal volumes of 50 µl, 150 µl and 250 µl. Rice. Figure 8. HPLC UV Detector Signal Offset versus Time Plot for Case 1 (Acetonitrile and Ammonium Acetate as indicators) showing the mixing of solvent without static mixer, the new series of Mott static mixers, and three conventional mixers (actual data offset is 0.1 ( without mixer), 0.32, 0.48, 0.6, 0.7, 0.8, 0.9 mA respectively for better display effect). The percentage reduction of the base sine wave is calculated by the ratio of the amplitude of the sine wave to the amplitude without the mixer installed. The measured sine wave attenuation percentages for Cases 1 and 2 are listed in Table 2, along with the internal volumes of a new static mixer and seven standard mixers commonly used in the industry. The data in Figures 8 and 9, as well as the calculations presented in Table 2, show that the Mott Static Mixer can provide up to 98.1% sine wave attenuation, far exceeding the performance of a conventional HPLC mixer under these test conditions. Figure 9. HPLC UV detector signal offset versus time plot for case 2 (methanol and acetone as tracers) showing no static mixer (combined), a new series of Mott static mixers and two conventional mixers (actual data offsets are 0, 11 (without mixer. ), 0.22, 0.3, 0.35 mA and for better display). Seven commonly used mixers in the industry were also evaluated. These include mixers with three different internal volumes from company A (designated Mixer A1, A2 and A3) and company B (designated Mixer B1, B2 and B3). Company C only rated one size.
Table 2. Static Mixer Stirring Characteristics and Internal Volume Static Mixer Case 1 Sinusoidal Recovery: Acetonitrile Test (Efficiency) Case 2 Sinusoidal Recovery: Methanol Water Test (Efficiency) Internal Volume (µl) No Mixer – - 0 Mott 30 65% 67.2% 30 Mott 60 92.2% 91.3% 60 Mott 90 98.1% 97.5% 90 Mixer A1 66.4% 73.7% 50 Mixer A2 89.8% 91.6% 150 Mixer A3 92.2% 94.5% 250 Mixer B1 44.8% 45.7% 9 35 Mixer B2 845.% 96.2% 370 Mixer C 97.2% 97.4% 250
Analysis of the results in Figure 8 and Table 2 shows that the 30 µl Mott static mixer has the same mixing efficiency as the A1 mixer, i.e. 50 µl, however, the 30 µl Mott has 30% less internal volume. When comparing the 60 µl Mott mixer with the 150 µl internal volume A2 mixer, there was a slight improvement in mixing efficiency of 92% versus 89%, but more importantly, this higher level of mixing was achieved at 1/3 of the mixer volume. similar mixer A2. The performance of the 90 µl Mott mixer followed the same trend as the A3 mixer with an internal volume of 250 µl. Improvements in mixing performance of 98% and 92% were also observed with a 3-fold reduction in internal volume. Similar results and comparisons were obtained for mixers B and C. As a result, the new series of static mixers Mott PerfectPeakTM provides higher mixing efficiency than comparable competitor mixers, but with less internal volume, providing better background noise and a better signal-to-noise ratio, better sensitivity Analyte, peak shape and peak resolution. Similar trends in mixing efficiency were observed in both Case 1 and Case 2 studies. For Case 2, tests were performed using (methanol and acetone as indicators) to compare mixing efficiency of 60 ml Mott, a comparable mixer A1 (internal volume 50 µl) and a comparable mixer B1 (internal volume 35 µl). , performance was poor without a mixer installed, but it was used for baseline analysis. The 60 ml Mott mixer proved to be the best mixer in the test group, providing a 90% increase in mixing efficiency. A comparable Mixer A1 saw a 75% improvement in mixing efficiency followed by a 45% improvement in a comparable B1 mixer. A basic sine wave reduction test with flow rate was carried out on a series of mixers under the same conditions as the sine curve test in Case 1, with only the flow rate changed. The data showed that in the range of flow rates from 0.25 to 1 ml/min, the initial decrease in the sine wave remained relatively constant for all three mixer volumes. For the two smaller volume mixers, there is a slight increase in sinusoidal contraction as the flow rate decreases, which is expected due to the increased residence time of the solvent in the mixer, allowing for increased diffusion mixing. The subtraction of the sine wave is expected to increase as the flow decreases further. However, for the largest mixer volume with the highest sine wave base attenuation, the sine wave base attenuation remained virtually unchanged (within the range of experimental uncertainty), with values ranging from 95% to 98%. Rice. 10. Basic attenuation of a sine wave versus flow rate in case 1. The test was carried out under conditions similar to the sine test with variable flow rate, injecting 80% of an 80/20 mixture of acetonitrile and water and 20% of 20 mM ammonium acetate.
The newly developed range of patented PerfectPeakTM inline static mixers with three internal volumes: 30 µl, 60 µl and 90 µl covers the volume and mixing performance range required for most HPLC analyzes requiring improved mixing and low dispersion floors. The new static mixer achieves this by using new 3D printing technology to create a unique 3D structure that provides improved hydrodynamic static mixing with the highest percentage reduction in base noise per unit volume of internal mixture. Using 1/3 of the internal volume of a conventional mixer reduces base noise by 98%. Such mixers consist of interconnected three-dimensional flow channels with different cross-sectional areas and different path lengths as the liquid crosses complex geometric barriers inside. The new family of static mixers provide improved performance over competitive mixers, but with less internal volume, resulting in better signal-to-noise ratio and lower quantitation limits, as well as improved peak shape, efficiency and resolution for higher sensitivity.
In this issue Chromatography – Environmentally friendly RP-HPLC – Use of core-shell chromatography to replace acetonitrile with isopropanol in analysis and purification – New gas chromatograph for…
Business Center International Labmate Limited Oak Court Sandridge Park, Porters Wood St Albans Hertfordshire AL3 6PH United Kingdom