Embed Size (px)
Citation preview
Lab on a Chip
Publ
ishe
d on
28
Nov
embe
r 20
13. D
ownl
oade
d by
Uni
vers
ity o
f O
ntar
io I
nstit
ute
of T
echn
olog
y on
22/
01/2
014
14:5
4:50
.
TECHNICAL INNOVATION View Article OnlineView Journal | View Issuea Faculty of Engineering and Applied Science, University of Ontario Institute of
Technology, 2000 Simcoe Street North, Oshawa, Ontario, Canada L1H 7K4.
E-mail: [email protected]; Tel: +1 905 721 8668bDepartment of Mechanical and Industrial Engineering and Institute for
Sustainable Energy, University of Toronto, 5 King's College Road, Toronto,
Ontario, Canada M5S 3G8. E-mail: [email protected]; Tel: +1 416 978 1623
† Electronic supplementary information (ESI) available: 1 video and a document.See DOI: 10.1039/c3lc51255j
Lab CThis journal is © The Royal Society of Chemistry 2014
Cite this: Lab Chip, 2014, 14, 681
Received 8th November 2013,Accepted 28th November 2013
DOI: 10.1039/c3lc51255j
www.rsc.org/loc
Out-of-plane ion concentration polarization forscalable water desalination†
Brendan D. MacDonald,*a Max M. Gong,b Pei Zhangb and David Sintonb
We present a scalable, out-of-plane desalination approach using ion concentration polarization.
A depletion boundary separates salt ions and purified water into distinct vertical layers. The out-of-plane
design enables multiplexing in three dimensions, providing the functional density required for
practical application. For membrane widths of 125–200 μm, and applied voltage of 5 V, the energy
requirement is 4.6 Wh L−1 for 20 mM solution, and 13.8 Wh L−1 for 200 mM. Energy efficiency is found
to be insensitive to flow rate as the depletion boundary adjusts to yield a commensurate volume of
purified water. Scaled-up devices are presented, which have a 3-fold improvement in functional density
over planar systems.
Introduction
Access to fresh water is crucial for human health; however,there is a limited supply. Reduced precipitation andvanishing glaciers currently threaten the water supply ofmore than one-sixth of the Earth's population.1,2 Many ofthe countries facing declining fresh water sources have readyaccess to seawater, which accounts for 97.5% of the wateron the Earth.3
Desalination is a capital- and energy-intensive process,which makes it challenging to implement in developingregions.3 Current technology, such as reverse osmosis, operatesmost efficiently at large scales, necessitating large centralizedfacilities. In addition to energy and cost aspects, centralizedfacilities have limited reach. In Vietnam for instance, ~70%of the population lives in rural communities (~60 millionpeople).4 A smaller, scalable desalination technology withhigh energy efficiency could provide fresh water to ruralpopulations on an individual, family, or village scale. A promis-ing technology with the capacity to meet these criteria is desali-nation by ion concentration polarization (ICP).5–7
ICP is a phenomenon that arises at the junction ofmicrofluidic and ion selective nanofluidic channels when anelectric field is applied.8 A depletion zone forms at the junc-tion, which repels charged particles.9 ICP has been largely
applied as a method for concentrating analytes.10–14
Kim et al. demonstrated application of ICP for the inverseprocess – concentrating ions from salt water with pure wateras the output.5,6 The desalination device was a planarmicrofluidic device with a Nafion® nanoporous membrane.The energy efficiency of their device was comparable to large-scale reverse osmosis plants in spite of being a microscaledevice. Further studies have explored planar nanoscale path-ways,12 slices in PDMS filled with a nanoporous membrane,15
and triangular nanochannels along the edge of planarnanoporous membranes.16
All previous work on desalination using ICP has involvedplanar designs with the purified water, concentrated stream,and ground channel located in the same layer. While planardevices are well suited to microfabrication, visualization, andproof-of-concept, they are not suited to scale-up. Traditionalplanar lab-on-a-chip designs have been effective for analysis,but planar designs are not volumetrically efficient when afluid product is of interest. In short, traditional chips aremostly plastic, whereas successful process technologies aremostly fluid. In order to be effective in applications wherethere is a fluid-product, as opposed to the more commoninformation-product, a re-imagining of the planar design isrequired.17 Transitioning lab-on-a-chip approaches, such asdesalination by ICP, to practical technologies, will requireeffective utilization of the third dimension.
In this work we present a scalable desalination deviceusing ICP in the out-of-plane direction. This approach gener-ates scalable, cost-effective desalination units that can becustomized for desired water output levels. We present theresults of experiments with a single level unit cell to quantifyunit operation and assess the impact of the geometry of thevertical nanoporous membrane and the effect of varying flow
hip, 2014, 14, 681–685 | 681
Fig. 2 (a) Schematic of the microfluidic device used for thecharacterization experiments to enable visualization of the depletionregion. A single channel was fabricated on the bottom layer so aninverted microscope could capture the depletion effect. (b) Schematicof the microfluidic device used to illustrate out-of-plane ICP depletionand deflection of ions, and demonstrate the scalability.
Lab on a ChipTechnical Innovation
Publ
ishe
d on
28
Nov
embe
r 20
13. D
ownl
oade
d by
Uni
vers
ity o
f O
ntar
io I
nstit
ute
of T
echn
olog
y on
22/
01/2
014
14:5
4:50
. View Article Online
rates. We present two scaled-up desalination devices, onewith an inlet channel width of 2 cm and another withthree vertically stacked units to demonstrate the scalabilityof the design.
Experimental
The out-of-plane ICP process is illustrated in Fig. 1. Electricfields were applied through positive electrodes connected tothe source water stream and grounded electrodes connectedto a channel located above and perpendicular to the sourcechannel. A vertical nanoporous membrane was used as apathway for ion transport between the microfluidic layers.When an electric field was applied, the cation selectivity ofthe nanoporous membrane generated a depletion effect onthe anodic side of the membrane, which redirected ions intothe lower channel and prevented ions from travelling into thetop channel, as shown in Fig. 1. As a result, the top channelcontained a purified water stream and the lower channelcontained a concentrated stream.
Experiments to characterize the effect of the size of thenanoporous membrane and the flow rate were performedusing the device shown in Fig. 2a. To visualize the depletioneffect, the channel below the nanoporous membrane wasconstructed as a single level with the membrane openingcovering a portion of the channel, so that ions were deflectedaround it and viewed with an inverted microscope. Theexperiments demonstrating the stackable functionality ofthe design were performed using the device shown in Fig. 2b.The device incorporates a multilayer source channel, asshown in Fig. 1, and the purified channel is deviated to allowfor visualization of ion removal.
The vertical nanoporous membrane was fabricated byslicing through a piece of silicone rubber with a CO2 laser
Fig. 1 A schematic representation of the microfluidic device at thenanoporous membrane where ICP generates a depletion boundary.Charged species are deflected downwards by the depletion boundaryinto a separate layer, purifying the top stream. The purified,concentrated, and ground channels are located on distinct layers,which enables stacking for increased functional density.
682 | Lab Chip, 2014, 14, 681–685
(see Fig. S1†), and filling with a solution of 5% Nafion® 117.The microfluidic device was fabricated by micromachiningthe channels and reservoirs into PMMA using a CO2 laser.Producing out-of-plane designs with the CO2 laser permits awide range of dimensions, thus devices can be produced forvarious capacity demands. More details regarding materialsand device fabrication are in the ESI.†
Results and discussionEffect of nanoporous membrane geometry
The dimensions of the nanoporous membrane constrain thetransport of cations through it, which can influence the sizeand effectiveness of the depletion boundary formed duringICP. There is a trade-off between the enhanced transport ofions through a larger nanoporous layer and the desire tokeep the depletion effect localized and focused by using asmaller nanoporous membrane. The nanoporous membraneswere cut to a length slightly smaller than the channel size(~1.8 mm), and the width values were varied. The experi-ments were run using the layout shown in Fig. 2a with achannel depth of 200 μm and width of 2.0 mm, a flow rate of
This journal is © The Royal Society of Chemistry 2014
Lab on a Chip Technical Innovation
Publ
ishe
d on
28
Nov
embe
r 20
13. D
ownl
oade
d by
Uni
vers
ity o
f O
ntar
io I
nstit
ute
of T
echn
olog
y on
22/
01/2
014
14:5
4:50
. View Article Online
0.5 μL min−1, and an applied voltage of 5 V. Salt solutions of20 mM and 200 mM were used for the experiments.
The depth of the depletion boundary was measured usingthe intensity of the fluorescence signal to determine thepercentage of the channel depth that was devoid of ions(directly downstream of the membrane), since ions wereprimarily redirected to the sides or halted in front of thedepletion boundary. Larger nanoporous membranes resultedin deeper depletion boundaries as shown in Fig. 3a. Theincrease in the depletion boundary size declined as themembrane width was increased, until the effect levelled off at~125 μm. In these experiments the voltage was set at 5 V andthe current was free to fluctuate. The current remainedrelatively steady for the various widths with an average valueof 27 μA for the 20 mM solution and 80 μA for 200 mM. Sincelarger membranes enable more ion transport they offer alower electrical resistance in the circuit. With lower resistancein the membrane, the depletion boundary grows in step,increasing the resistance, leading to a relatively constantcurrent (and thus constant power) over a range of membranewidths. These results indicate that the membrane area shouldbe made sufficiently large to achieve the minimum desired
Fig. 3 Experimental results for various nanoporous membrane widthsat a flow rate of 0.5 μL min−1 and applied voltage of 5 V. (a) Iondepletion percentage is plotted versus nanoporous membrane width.The depletion percentage correlates to the depletion boundary depth,which is shown on the secondary y-axis. The results indicate thatlarger nanoporous membranes generate larger depletion boundaries.(b) Energy required per litre output is plotted versus nanoporousmembrane width.
This journal is © The Royal Society of Chemistry 2014
depletion boundary size. In the stackable configuration(Fig. 2b) the desired size is the full extent of the opening tothe purified water channel (Fig. 1).
In terms of energy usage per volume of purified water,larger nanoporous membranes are more efficient, as shownin Fig. 3b. This is due to larger depletion boundaries resultingfrom the lower electrical resistance of larger nanoporousmembranes, while the current (and power) remain relativelyconstant. The improvement in energy efficiency levelled off at~125 μm for both salinities tested. Beyond that point therequired energy was 4.6 Wh L−1 for the 20 mM solution, and13.8 Wh L−1 for 200 mM. These results demonstrate theefficacy of the vertical nanoporous membrane (out-of-plane)approach and the efficiency benefit from larger nanoporousmembranes.
Effect of flow rate
The flow rate through the channel determines how many ionsare being transported to the depletion boundary andconsequently how many ions are deflected. Experiments wererun using the same device as the nanoporous membranegeometry experiments, with a fixed membrane width of 90 μmand an applied voltage of 8 V. The nanoporous membranewas cut to a length shorter than the channel width (1.0 mm)to allow ions to be deflected to the side of the depletionboundary at higher flow rates.
The depth of the depletion boundary decreased as the flowrate was increased, as shown in Fig. 4a. The decrease in depthwas approximately linear with increasing flow rate such thatat each flow rate approximately the same amount of ions werebeing deflected around the depletion boundary. This effectresulted in consistent outgoing flow rates of the purifiedwater for all of the incoming flow rate values tested. Theelectrical power was also relatively consistent at all flow rateswith an average current value of 54 μA for the 20 mM solutionand 89 μA for 200 mM. As a result of the consistent purifiedflow rates and electrical power values, we see in Fig. 4b thatthe required energy was the same (within the error) for allflow rates. In summary, these results indicate that varying theflow rate has a negligible impact on the energy efficiencysince the depletion boundary deflects the same amount ofions; therefore generating the same amount of outgoingpurified water regardless of the incoming flow rate. In afunctioning device, the critical aspect is that the depletionboundary is sized to cover the full extent of the purified waterchannel (Fig. 1), by tuning flow rate, and/or membrane width.
Out-of-plane desalination device
To demonstrate out-of-plane ICP in a functional desalinationdevice we ran experiments with the device shown in Fig. 2busing a flow rate of 1.0 μL min−1. The channel dimensionsare listed in the ESI.† The width of the nanoporous mem-brane in these experiments was 90 μm, and the applied volt-age was 20 V. We used 500 mM salt solution for theseexperiments since it is a reasonable analog to seawater. As
Lab Chip, 2014, 14, 681–685 | 683
Fig. 4 Experimental results for various flow rates with a nanoporousmembrane width of 90 μm and applied voltage of 8 V. (a) Iondepletion percentage and depletion boundary depth is plotted versusflow rate. The results indicate that the depletion boundary depthdecreased in a linear manner as the flow rate was increased. (b) Energyrequired per litre output is plotted versus flow rate. The energyefficiency was consistent for incoming flow rates due to the consistentflow rate of purified water.
Fig. 5 Experimental demonstration of out-of-plane desalination.(a) Image of the experiment taken before the voltage was applied and(b) 20 seconds after 20 V was applied. The flow rate was 1.0 μL min−1.The source water had a salt concentration of 500 mM and was sepa-rated into purified and concentrated layers (video available in the ESI†).
Lab on a ChipTechnical Innovation
Publ
ishe
d on
28
Nov
embe
r 20
13. D
ownl
oade
d by
Uni
vers
ity o
f O
ntar
io I
nstit
ute
of T
echn
olog
y on
22/
01/2
014
14:5
4:50
. View Article Online
noted above, the multilayered nature of this device, whilenecessary from a volumetric efficiency perspective, inherentlycomplicates visualization. In order to accommodate quantifi-cation of the purification process, the channel from the upperlayer was deviated in-plane, as seen in Fig. 5 (this would notbe required or desired in an integrated device). As shown inFig. 5, the upper-level stream was visibly devoid of dye, andquantitative fluorescence measurements indicated ~95%purity. A video is available in the ESI.†
In order to demonstrate the scalability of the device weused the layout in Fig. 2b with two variations: one with awider inlet channel (and correspondingly longer nanoporousmembrane) and another with stacked purification levels. Wefabricated a device with an inlet channel width of 2 cm, useda flow rate of 20 μL min−1, an applied voltage of 20 V, andachieved a purification of ~90% for 500 mM salt solution(device is shown in Fig. S2†). We also produced double andtriple layer devices, with the same dimensions as the devicein Fig. 5, to demonstrate stacking. Visualization was possiblefor the bottom level and successful ion deflection was visuallyconfirmed for both devices. Successful operation of all thelevels was confirmed by comparing the electrical current,
684 | Lab Chip, 2014, 14, 681–685
since an approximately linear increase was measured withvalues of 60 μA, 110 μA, and 190 μA, for single, double, andtriple level devices respectively.
An advantage of the vertical nanoporous membrane inthe out-of-plane design is that it permits an expansion ofthe channel width parallel to the length of the membranewithout restricting the function of ICP, as demonstrated bythe device with a channel width of 2 cm. In previousplanar devices, where ions were transported through thenanoporous membrane to a microfluidic channel beside thepurification layer, there was a limit to the channel widthbased on the size of the depletion boundary. Improvementsin functional density depend on the operating conditionsand scale of the design. For the operating conditions andscale of the triple layer device, we calculated that ~45% ofthe unit device volume is fluid, whereas previous planardevices were ~15% fluid. The out-of-plane device has afunctional density that is three times greater; however, theprevious devices were not optimized for functionaldensity. We estimate ~50% greater functional density forout-of-plane devices if the planar devices are optimized. Insummary, the scalability of the device was demonstrated andthe out-of-plane design enabled significantly greaterfunctional density than planar devices.
Conclusions
In this work, we demonstrated scalable, out-of-planedesalination using ICP. Experimental results established thatusing larger nanoporous membranes for ICP yields largerdepletion boundaries and higher energy efficiency. For larger
This journal is © The Royal Society of Chemistry 2014
Lab on a Chip Technical Innovation
Publ
ishe
d on
28
Nov
embe
r 20
13. D
ownl
oade
d by
Uni
vers
ity o
f O
ntar
io I
nstit
ute
of T
echn
olog
y on
22/
01/2
014
14:5
4:50
. View Article Online
membrane widths (125–200 μm), at an applied voltage of 5 V,the required energy was 4.6 Wh L−1 for the 20 mM solution,and 13.8 Wh L−1 for 200 mM. Flow rate variation was found tohave a negligible impact on energy efficiency since the size ofthe depletion boundary adjusted to yield a consistent amountof purified water for a given applied voltage. A device with a2 cm inlet channel and a three level device were presented todemonstrate scalability using 500 mM salt solution, and werefound to have 3 times the functional density of planar designs.The out-of-plane design presents opportunities for practicallow cost desalination devices.
Acknowledgements
The authors would like to gratefully acknowledge fundingfrom Grand Challenges Canada through Stars in GlobalHealth, funding from CFI, and ongoing support from NSERC.
Notes and references
1 T. P. Barnett, J. C. Adam and D. P. Lettenmaier, Nature,
2005, 438, 303–309.2 J. Xu, R. E. Grumbine, A. Shrestha, M. Eriksson, X. Yang,
Y. U. N. Wang and A. Wilkes, Conserv. Biol., 2009, 23,520–530.3 M. A. Shannon, P. W. Bohn, M. Elimelech, J. G. Georgiadis,
B. J. Marinas and A. M. Mayes, Nature, 2008, 452, 301–310.This journal is © The Royal Society of Chemistry 2014
4 General Statistics Office of Vietnam, [online], available from
http://www.gso.gov.vn/default_en.aspx?tabid=467&idmid=3.5 S. J. Kim, S. H. Ko, K. H. Kang and J. Han, Nat.
Nanotechnol., 2010, 5, 297–301.6 S. J. Kim, S. H. Ko, K. H. Kang and J. Han, Nat.
Nanotechnol., 2013, 8, 609.7 E. Choi, K. Kwon, S. J. Lee, D. Kim and J. Park, Proceedings
of IEEE 25th International Conference on Micro ElectroMechanical Systems (MEMS), Paris, 2012.8 A. Höltzel and U. Tallarek, J. Sep. Sci., 2007, 30, 1398–1419.
9 Q. S. Pu, J. S. Yun, H. Temkin and S. R. Liu, Nano Lett.,2004, 4, 1099–1103.10 Y. C. Wang, A. L. Stevens and J. Y. Han, Anal. Chem.,
2005, 77, 4293–4299.11 J. H. Lee, Y. A. Song, S. R. Tannenbaum and J. Han, Anal.
Chem., 2008, 80, 3198–3204.12 J. H. Lee, Y. A. Song and J. Y. Han, Lab Chip, 2008, 8,
596–601.13 B. Scarff, C. Escobedo and D. Sinton, Lab Chip, 2011, 11,
1102–1109.14 S. H. Ko, Y.-A. Song, S. J. Kim, M. Kim, J. Han and
K. H. Kang, Lab Chip, 2012, 12, 4472–4482.15 S. J. Kim and J. Han, Anal. Chem., 2008, 80, 3507–3511.
16 B. Kim, J. Heo, H. J. Kwon, S. J. Cho, J. Han, S. J. Kim andG. Lim, ACS Nano, 2012, 7, 740–747.17 D. Erickson, D. Sinton and D. Psaltis, Nat. Photonics,
2011, 5, 583–590.
Lab Chip, 2014, 14, 681–685 | 685
