University of Groningen
Microfluidic tools for multidimensional liquid chromatography Ianovska, Margaryta
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date: 2018
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Ianovska, M. (2018). Microfluidic tools for multidimensional liquid chromatography. University of Groningen.
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
Chapter VI
General discussion,
conclusions and future
General discussion, conclusions and future perspectives
179
General discussion, conclusions and future perspectives
In this thesis, we have demonstrated the development and application of the chaotic microfluidic mixer for improving conventional analytical technique such as two-dimensional liquid chromatography (2D-LC). Without doubts, the research described here provides intriguing possibilities for the application of the microfluidic technology. While many miniaturized lab-on-a-chip systems trying to replace large laboratory equipment (which in many cases is not achievable task), this project was focusing on using advantages of the miniaturized devices (such as small volume) to be connected on-line with conventional equipment. The research showed in this thesis arose as a need to find more convenient approaches to improve the performance of the 2D-LC technique and to bring it to a different level, where analysis of tens thousand and more components in one run become possible.
The research, presented in this thesis, covers many different facets of modern technology. This work has been done in the interface between two rapidly developing fields – multidimensional LC and microfluidics. On one hand we talk about huge complicated equipment, composed of 2 chromatographic systems coupled through the intricate interface. On the other hand, we have a small 5-cm long chip, with, nevertheless, a potential to solve one of the major problems that exist in 2D-LC separation. The problem that we are talking about here is a solvent incompatibility between dimensions and we propose a solution to it such as the utilization of the mixing device in the interface.
As was discussed in Chapter 1 and Chapter 3, the mixing device in the interface between two columns in 2D-LC must satisfy three strict conditions: it should provide fast mixing in-line at different ratios in the wide range of flow rates (compatible with typical flow rates used in 2D-LC); have a small volume (to not contribute to the extra column-band broadening); and must be able to withstand brief pressure pulses up to a few hundred bar (due to its connection to switching valves which operated under the pressure from pumps). Therefore, an idea to develop and use small-volume microfluidic mixers that are able rapidly mix solutions at different flow rates is justified.
It is important to keep in mind, that microfluidic devices, on the other hand, has a very important and disadvantageous in our situation feature such as an existence of a laminar flow, which does not allow two streams to be mixed under normal consequences. Therefore, our first
research goal was to develop a fast micromixer that works under the wide range of flow rates Dis
cu
ss
Microfluidic Tools for Multidimensional Liquid Chromatography
180
and able to mix solvents with different viscosity (which is often a case in LC). Based on the overview we presented in Chapter 2, it was clear that the mixer with herringbone grooves meets the required conditions. In Chapter 3 we proved this concept. Besides, we believe that the wide overview of the existing micromixers based on chaotic advection and the approach for choosing the appropriate type for a particular application presented in Chapter 2, combined with the fabrication protocol described in Chapter 3 provide a good base for the end-users who are interested in using microfluidic mixers in their own research.
As Robert M. Pirsig said, “Technology presumes there is just one right way to do things and there never is”. Seeing the presented research from this perspective, the fabricated devices were designed with the idea to provide competition for the existing mixers – T-piece and S-mixers1 – that are currently used for mixing in the conventional LC systems. The developed by us micromixer possess better characteristics in terms of volume and efficiency of mixing under lower flow rates. As we showed later in Chapter 5, the developed COC-micromixer (with the volume of 4.65 µL, M-mixer) shows similar dispersion profile to T-piece and the S-mixer with the volume to 10 µL. M-mixer is clearly outperforms S-mixer with 150 µL volume in both dispersion profile and the real application. However, in the real separation T-piece and S-mixer give similar performance while M-mixer give some time-shift and small tailing. Even if the mixer will show better performance than commercial mixers at lower flow rate (for which it was designed), the overall dispersion due to the relatively large volume of the system (loops and tubing) will vanish the benefit of having a small volume of the M-mixer. We suggest that the benefit of the developed micromixer would be seen to the full extend if other component have also smaller volume and integrated on the same chip. In other words, if the modulator and a second-dimension column are also integrated on the same micromachined device. This will dramatically decreased not only the dispersion of the system (due to the absence of extra tubing for connecting different components), but also reduce the time of the second-dimension separation. Figure 1 represents the situation when a column packed with a stationary phase (particles or monoliths) is used instead of loops in the modulator. In such system, the effluent coming from the first dimension will be mixed with the modifying solvent in order to trap analytes on the pre-concentration column, followed by elution of the analytes after switching the solvent with higher eluting strength. In this case, this will create so-called trap column that provide an extra pre-concentration of analytes before entering the second column and it would provide an active modulation in the LC×LC interface. Afterwards, the analytes will proceed to
General discussion, conclusions and future perspectives
181
the second-dimension where the effluent will be mixed with the 2D mobile phase, after which the second-dimension separation will take place in the channel packed with a stationary phase. We believe that developing such state-of-the-art device one should be concerned about several aspects. The uniform packing of the microfluidic channel with particles can be a challenging task, it require utilization of special frits and high pressure for packaging. The fabrication of frits is problematic within a microfluidic channel and formation of bubbles and band broadening is frequently observed. In case of monoliths, during the polymerization process the gaps between channel walls and the monolithic phase can form. Besides, the monoliths can dislodge under applied pressure. In both cases, the utilization of the channel with stationary phase will create additional backpressure that should be taking into account designing the whole separation system.
Figure 1. A micromachined device with integrated modulator with a mixer and pre-concentration column and
mixer for modifying the mobile phase followed by the second-dimension separation column. Additional pump assists in the absorption/desorption process from the pre-concentration column.
Another example of the device that our mixer could compete with is a commercially available Jet Weaver mixer,2 developed by Agilent and incorporated into the HPLC pumping system (1290 Infinity Binary pump). That is a perfect example of the chip-based microfluidic device that have already successfully entered the world of “macro-equipment”. As discussed in
Chapter 3, this mixer employs the split-and-recombine mixing principle using fro this a network
of multi-layer microfluidic channels. The minimum volume of this device is 35 µL. In our approach we use different mixing mechanism – chaotic advection, which was proved to be more efficient mixing mechanism that allows to obtain mixing using devices with smaller volume, as discussed in Chapter 2. Besides, our mixer can be separately connected to any LC or MS equipment, while Jet Weaver mixer is not applicable outside LC pumping system.
Dis
cu
ss
io
Microfluidic Tools for Multidimensional Liquid Chromatography
182
Another important requirement that was set for the mixing device is its pressure-resistance. The originally reported design by Stroock et al.3 had a very small inner volume (around 1 µL). In order to maintain pressure drops below 1 bar, we designed micromixers with increased dimensions (obtaining the inner volumes of 1.6 μL and 2.2 μL). However, the dimensions used by Stroock et al.3 can not simply be multiplied by a constant to achieve mixers with bigger volumes exhibiting the same mixing efficiency. Therefore, a lot of work was put into optimization of channel and groove parameters based on a previously described numerical study.4
As was shown in Chapter 3, our initial devices were fabricated in PDMF using soft-lithography. While it was sufficient for the prove of concept, these devices didn’t withstand more than 10 bar of pressure. Starting to search a new fabrication method in order to obtain pressure-resistant chips revealed that the initial chip features around 50 µm is on the edge of the resolution of all accessible to us fabrication techniques. Therefore, in order to fabricate a pressure-resistant chip, we had to modify and optimize the geometrical features inside the channel again. Using the same approach described in Chapter 3, we obtained chips with increased dimensions and the minimum features of 100 µm. Thought it is slightly increased the inner volume of the device from 1.6 to 3.6 µL (Chapter 4), it allowed us to consider several fabrication methods.
In Chapter 4 and Chapter 5 we were exploring two relatively new techniques, such as Selective Laser-Induced Etching (SLE) and micromilling. For the work descried in Chapter 4, a solid block of fused silica was used to create a chip using the laser modification of the fused silica with the following etching of modified material. Because the whole procedure was performed in the solid piece of fused silica, the bounding step (bonding of two parts of the chip) was eliminated. We aimed to fabricate a 30-mm-long microfluidic mixing channel with herringbone structures. Unfortunately, due to such channel length, the etching time required for removing the modified material was also too long and we didn’t manage to obtain the equal cross-section of the channel. Besides, the herringbone grooves were observed only in the middle part of the channel, which was not sufficient to obtain a good working mixing device. However, in terms of pressure-resistance, the obtained fused-silica chips can withstand pressure of 85 bar, which make them applicable in the interface of multidimensional liquid chromatography as a separate mixing device.
General discussion, conclusions and future perspectives
183
We tried to solve these fabrication issues accounting for difference in etching rate in different device regions (more etching from the channel side and less in the middle). This brought us to the idea of an adjusted “gradient” design with different channel width/depth and groove dimensions at the beginning/end and in the middle of the channel (depending on the time that a particular region stays in contact with the aching solution). The next chip generation with such adjusted design showed more equal cross section and much larger regions with good-resolved herringbone grooves, which lead to a better degree of mixing. Unfortunately, a complete mixing by the end of the channel was still not obtained. In the future, it would be interesting to optimize design further and obtain a chip with good-resolved grooves along the whole channel.
Nevertheless, to the best of our knowledge, the fabrication of the fused silica chips with such length with the complex features inside using the SLE technique has never been explored before. This makes our work a pioneering in this area of research and makes an interesting topic for the future.
Due to the failure of the fabrication approach exploited in Chapter 4. we have to search for another option for the fabrication of pressure-resistant chip. We chose micromilling technique as a compromise between speed of transferring the sketch into the actual device (it is a direct fabrication method), its accessibility and appropriate resolution (around 100 µm). The choice of material was also an important aspect. We decided to use cyclic-olefin copolymer (COC), due to it is rigidity and compatibility with solvents used in LC applications (acetonitrile, methanol, etc.). The fabrication method, as described in Chapter 5, consisted of several steps, which, due to the limited time, were strategically planned. The first part was milling in COC substrate. It required 5 hour in total to mill one bottom part of the chip with a channel and herringbone structures. The top part was cut from the sicker piece of COC with the drilled conical access holes. The next step was the bonding of two parts, which was done using the solvent-vapour-assisted bonding approach that was used in the group of Eeltink.5 Due to the mechanical interlocking of polymer chains between two surfaces under the cyclohexane vapor, exceptionally strong bond between two parts was achieved.6 Without doubts, this approach proved to be very useful in our application, but can be easily applied for the fabrication of any pressure-resistance devices. The last step of the fabrication process was development of a specially designed holder to assist in the pressure-resistance of the device and to connect the chip to the equipment. The holder consists of two metal parts and the chip was clamped between
Dis
cu
ss
io
Microfluidic Tools for Multidimensional Liquid Chromatography
184
them. The top metal part had an access wholes that had a standard 10-32 thread for standard HPLC connectors. The final assembly was able to withstand pressures up to 200 bar.
Due to the standard 10-32 thread in the metal top part and the conical parts in the top COC plate, the holder also served as a low-dead-volume micro-to-macro interface to directly coupled the chip to 2D-LC system. This approach opens an attractive perspective for the future implementation of any microfluidic component into conventional instrumentation, even at high pressures.
The developed COC chip was implemented in a HILIC×RPLC system for successful separation and identification of various oligomeric series in Polyamide 46 samples. As was shown in Chapter 5, the microfluidic mixer successfully copes with the mobile-phase mismatch between two dimensions, mixing the effluent after the first column before transferring it to the second dimension. The proof of good mixing can be seen as the prevention of the break-through in the second dimension. We want to emphasize that even though the developed mixer did not show better performance than T-piece and 10-uL S-mixer, there is no need to underestimate the developed mixer. Having a very small volume, it proved to be fast, efficient and robust during a very large number of modulations. Unfortunately, the time of this project was limited and there was no opportunity to perform more detailed research and to compare thoroughly the behavior of different mixing units. It would be also interesting to investigate performance of the microfluidic mixer in the separation of different samples and therefore, at different flow rates.
In conclusion, the results of this thesis show a very good potential of combining the microfluidic technology and conventional analytical techniques. In our case, we aimed to improve the performance of multidimensional liquid chromatography using a microfluidic mixer. In this work, we provided a comprehensive methodology for choosing an appropriate micromixer and design adjustments for the need of particular application, its fabrication using several techniques, its integration into the conventional LC equipment and its successful application even at high pressures. We believe that the research described in this thesis gives a good basis for scientists, who has an interest in both microfluidics and multidimensional liquid chromatography for their own research.
General discussion, conclusions and future perspectives
185
References
1. Brochure provided by Analytical Scientific Instruments US, http://www.hplc-asi.com/static-mixers.
(2014).
2. Agilent 1290 Infinity LC System, Manual and Quick Reference, Agilent Technologies. (2012).
3. Stroock, A. D. et al. Chaotic mixer for microchannels. Science 295, 647–651 (2002).
4. Lynn, N. S. & Dandy, D. S. Geometrical optimization of helical flow in grooved micromixers. Lab Chip 7,
580–587 (2007).
5. Wouters, B. et al. Design of a microfluidic device for comprehensive spatial two-dimensional liquid
chromatography. J.Sep.Sci. 38, 11123–1129 (2015).
6. Tsao, C. W. & DeVoe, D. L. Bonding of thermoplastic polymer microfluidics. Microfluid. Nanofluidics 6,
1–16 (2009). Dis cu ss io n
