Electrofluidic Circuit-Based Microfluidic Viscometer ... - ACS Publications

Subscriber access provided by UNIV OF NEW ENGLAND ARMIDALE

Article

An Electrofluidic Circuit-Based Microfluidic Viscometer for Analy-sis of Newtonian and Non-Newtonian Liquids Under Different Temperatures Tse-Ang Lee, Wei-Hao Liao, Yi-Fan Wu, Yeng-Long Chen, and Yi-Chung Tung Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.7b04779 • Publication Date (Web): 02 Jan 2018 Downloaded from http://pubs.acs.org on January 3, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 17 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

An Electrofluidic Circuit-Based Microfluidic Viscometer for Analysis of Newtonian and Non-Newtonian Liquids Under Different Temperatures Tse-Ang Lee,1 Wei-Hao Liao,1 Yi-Fan Wu,2,3 Yeng-Long Chen,2,3 and Yi-Chung Tung1,4,* 1 2 3 4

Research Center for Applied Sciences, Academia Sinica, Taipei 11529, Taiwan Department of Chemical Engineering, National Tsing-Hua University, Hsinchu 30013, Taiwan Institute of Physics, Academia Sinica, Taipei 11529, Taiwan College of Engineering, Chang Gung University, Taoyuan 33302, Taiwan

ABSTRACT: This paper reports a microfluidic viscometer with an integrated pressure sensor based on electrofluidic circuits, which are electrical circuits constructed by ionic liquid-filled microfluidic channels. The electrofluidic circuit provides a pressuresensing scheme with great long-term and thermal stability. The viscosity of the tested fluidic sample is estimated by its flow resistance, which is a function of pressure drop, flow rate, and the geometry of the microfluidic channel. The viscometer can be exploited to measure viscosity of either Newtonian or non-Newtonian power-law fluid under various shear rates (3 to 500 1/s) and temperatures (4 to 70oC) with small sample volume (less than 400 µl). The developed sensor-integrated microfluidic viscometer is made of polydimethylsiloxane (PDMS) with transparent electrofluidic circuit, which makes it feasible to simultaneously image samples under tests. In addition, the entire device is disposable to prevent cross contamination between samples, which is desired for various chemical and biomedical applications. In the experiments, viscosities of Newtonian fluids, glycerol water solutions with different concentrations and mixture of pyrogallol and sodium hydroxide (NaOH), and non-Newtonian fluids, Xanthan gum solutions and human blood samples have been characterized. The results demonstrate that the developed microfluidic viscometer provides a convenient and useful platform for practical viscosity characterization of fluidic samples for a wide variety of applications.

INTRODUCTION Viscosity is one of the most essential fluidic properties regulating fluidic behaviors, and it is a crucial factor for many applications. For instance, in industries of food and chemical manufacturing, viscosity is an important factor in product quality control and formulation optimization. In biochemical research, important information on mechanisms and dynamics of a molecular process can be obtained by studying dependence of solvent viscosity1-3. In medical diagnosis and prognosis, increases in viscosity of blood predict clinical manifestations of atherothrombotic vascular disease4. Furthermore, elevated blood viscosity is associated with diseases such as hypertension5,6, infarction7,8, cognition decline9 and diabetes10,11. In addition, viscosities of other body fluids, such as saliva12, synovial fluid13 and urine14, can also be indicators to monitor the progression of various diseases. In these biomedical applications, the samples are often precious with limited volumes. However, conventional viscometers usually require large sample volume (in order of tens ml), and sample holders are not disposable that may potentially cause cross-contaminations between samples. In addition to the constraints, there are still disadvantages among existing viscometers, for example: tedious and time-consuming process, expensive instrumentation and requirement of professional operators. In recent decades, microfluidics has drawn increasing attentions in chemistry and biological analysis because of its short

reaction time, and small sample and reagent volume consumption. To overcome the aforementioned challenges, various microfluidic devices have been developed for viscosity measurements. For example, Srivastava and Burns develop a microfluidic device made of glass and silicon capable of measuring viscosities of both Newtonian and non-Newtonian fluids based on capillary pressure-driven laminar flows15. The viscosity is calculated by measuring the advancing speed of fluidic samples using a microscope with a camera. Han et al. reports a microfluidic capillary device with negative pumping technique using the similar working principle as the device built by Srivastava and Burns16. Guillot et al. design a microfluidic device with a T-junction. Two immiscible fluids are injected into the device at the upstream, and the viscosity is calculated by investigating interface of the two fluids17. Kang et al. inject a fluidic sample into a straight microfluidic channel utilizing a syringe pump and calculate its viscosity from sample flow rate and total pressure drop, which is measured from an on-pump pressure transducer18. Taking advantage of the similar operation principle, Pan and Paulo19 measure the pressure drop using a series of piezoresistive sensors during sample injection to calculate the viscosity of Newtonian and non-Newtonian fluids at relatively high shear rates. Instead of measuring the total pressure drop, Chevalier and Ayela use microfabricated strain gauges to measure the local pressure for viscosity estimation20.

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Microfluidic viscometers provide advantages over conventional ones; however, they still have several drawbacks retarding their practical usage. For the image-based viscometers, complex imaging setup, image analysis, and data processing are commonly required for the viscosity characterization, which make their measurement automation difficult15-17. In addition, the limited shear rate ranges, the limitation for Newtonian fluid samples, relatively large errors at low shear rate ranges are also problematic for accurate measurements15-17. For the pressure sensor-based viscometers, additional expensive pressure transducers hinder the devices from using on a disposable basis18-20. In addition, extra pressure drop induce from tedious interconnections also affects the measurement results. The complicated configuration of the devices and setups result in difficult imaging of the samples simultaneously. Furthermore, most of the viscometers, including both imagebased and sensor-based viscometers, suffer from temperature sensitivity leading to recalibration of the devices or even inaccurate results during the measurements. To overcome the challenges, we develop a disposable and optically transparent microfluidic viscometer in this paper. The device is made of an optically transparent elastomeric material, polydimethylsiloxane (PDMS), due to its desired advantages. The viscosity is estimated by measuring pressure built up at upstream during a sample fluid flowing through a microfluidic channel. The pressure sensor is constructed using electrofluidic circuits that previously developed in our lab21-24. Electrofluidic circuits are electrical circuits constructed by microfluidic channels filled with ionic liquid. Ionic liquid possesses several advantageous material properties, including electrical conductivity, low vapor pressure, and thermal stability. The electrofluidic pressure sensors can be directly fabricated within PDMS microfluidic devices, and have great long term and temperature stability. Therefore, the developed microfluidic viscometer is disposable to prevent crosscontamination between samples. The developed sensorintegrated microfluidic device can be used for Newtonian and non-Newtonian fluid viscosity measurement under different temperatures and shear rates. In this paper, theoretical models are derived to estimate performance of the developed microfluidic viscometer. In the experiments, viscosities of both Newtonian and nonNewtonian fluidic samples are tested using the device. In addition, a Newtonian fluid is characterized under different temperatures to investigate the temperature stability of the device. The results show high measurement accuracy and precision provided by the device, and the measured viscosities agree well with the reported values. Furthermore, viscosity and color variation of a mixture of chemicals undergoing a chemical reaction are simultaneously investigated for an extended period to confirm the feasibility of real-time imaging of samples during the sample characterization due to the optical transparency of the device. For practical applications, human blood sample is characterized using the device under two different temperatures. The experimental results demonstrate that the developed microfluidic viscometer offers a reliable device for viscosity characterization using small volume of samples without concerns of cross-contamination that are highly desired for various analysis and applications.

MATERIALS AND METHODS

Page 2 of 17

Device Design The microfluidic viscometer is constructed using an elastomeric material, PDMS, due to its great optical transparency and manufacturability. The device consists of a glass substrate and two PDMS layers, a bottom microfluidic channel layer and a top electrofluidic circuit layer, with designed microfluidic channel patterns. A 30 µm-thick PDMS membrane is sandwiched between the two PDMS layers as shown in Figure 1(a). The microfluidic channel is designed with a flow chamber with a larger channel width at upstream to reduce flow speed and create a region with relatively uniform hydrostatic pressure, and with a serpentine shaped channel at downstream serving as a flow resistance. The hydrostatic pressure at the upstream is built up during sample injection, and the pressure magnitude is related to the channel geometries, and the flow rate and viscosity of the injected fluid. With known channel geometries, shear rate under a specific flow rate can be estimated. As a result, the viscosity characteristics at various shear rates can be estimated by measuring the upstream hydrostatic pressure at different flow rates. For viscosity measurement, sample fluid is injected into the device with controlled flow rates. In order to precisely measure the hydrostatic pressure, a pressure sensor based on electrofluidic circuits is constructed on the top layer. The electrofluidic circuit is constructed using ionic liquid-filled microfluidic channels with different geometries21. Ionic liquid, salt in liquid state, is an electrically conductive fluid with many desired material properties, such as negligible vapor pressure25, electrical conductivity26,27, and thermal stability28. Taking advantage of the electrofluidic circuit concept, electrofluidic resistors can be made by filling ionic liquid in long and narrow channels. In addition, pressure-regulated variable resistors can be constructed by filling ionic liquid into channels with flexible channel walls. Once the walls are deformed by pressure, the electrical resistance of the resistor can be changed due to the variation of its cross-sectional geometry as shown in Figure 1(b). As a result, the magnitude of the applied pressure can be estimated by measuring the electrical property change. In the microfluidic viscometer, an electrofluidic Wheatstone bridge circuit with three constant resistors and one pressureregulated resistor is exploited as a pressure sensor. The pressure sensor based on ionic electrofluidic circuit possesses excellent sensing linearity, thermal stability and long-term stability21. The circuit architecture provides accurate measurement of resistance change with temperature compensation. A pressure transduction hole is made on top of the flow chamber at the bottom layer for viscosity characterization. The microfluidic viscometer device is fabricated by a well-developed soft lithography replica molding technique29. The detailed fabrication process is described in the Supporting Information. In the developed microfluidic viscometer, the ionic liquid used in the experiments does not mix with other chemicals; therefore, no addition purification process is required before disposal. In addition, recycle of ionic liquids from the device can also be easily achieved by pipetting the liquids out from the microfluidic channels. The disposability and recycling of ionic liquids have been extensively discussed in recent literatures30. The relative small ionic liquid volume and open interconnection (inlets and outlets) design make it simple to recycle ionic liquid, which greatly improve the entire device disposability.

2 ACS Paragon Plus Environment

Page 3 of 17 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Theoretical Derivation In order to estimate the sample fluid viscosity from the output voltage variation of the pressure sensor, a theoretical model based on solid mechanics and basic electrical circuit theory is derived. The output voltage shift of the pressure sensor is derived to be linearly proportional to the pressure at the pressure transduction hole:

∝ ∆

(1)

where Pc is the gauge pressure at the pressure transduction hole and Vm is the pressure sensor output voltage. The detail derivation is shown in the Supporting Information. Taking advantage of equivalent fluidic circuit theory31, the flow resistance measuring friction in the flowing fluid can be written as:

=

∝

∆

(2)

where Rflow is the flow resistance and Q is the flow rate. According to equation (2), the flow resistance is proportional to the pressure sensor sensitivity, which is defined as output voltage shift under a unit flow rate change. The sensitivity can be measured during the experiments by measuring output voltage under different flow rates. For Newtonian fluids, viscosity can be expressed as:

(3)

where τ is the shear stress, γ is the shear rate, and µ is the viscosity of the fluidic sample. The explicit relation between shear stress and shear rate in terms of pressure drop and flow rate for steady flow in a rectangular channel can be derived as32: ℎ

6#

ℎ

ℎ

2 +ℎ = ℎ2 1 + %∗

%∗ ) =

2

*1 + 1 − )

192 -5 )

∑∞2=1,3,5,…

2-

012ℎ

25

2

)

7

−1

(5)

Rearranging the equation (4), the flow resistance can be written as:

%89 =

(7)

where µ and Rflow are the viscosity and measured flow resistance of the fluid sample, respectively; µc and Rflow,c is the viscosity and measured flow resistance of another fluid with known viscosity. For non-Newtonian fluids, based on power-law model33, the viscosity is expressed as:

@ = A BCD

(8)

where η is the apparent viscosity, n (power-law index) and K are constants characterizing the fluid: n>1 for shear thickening, n=1 for Newtonian fluids, and nL

;?,L>L

∙

;?,

;?, X ∗

∙

YBZ ∗ [\ ∗

M ?N O N ∙ V∗ L P?QOR STU∗ Q L W ?N O N MR?QOR S∙70%), the maximum flow rate is set to be 5 µl/min with 1 µl/min increment to reduce the required response time between different flow rate levels and the duration at each flow rate level is extended to 100 seconds to ensure the steady state of the voltage shift is attained. The experiments are performed in an environment with constant temperature of 23oC. Since the viscosity of glycerol solution varies with temperature, the experiment is also conducted under a broad range of temperature (4oC to 70oC) by placing the entire device into the refrigerator, including the syringe pump, or on a hotplate using the 60% glycerol solution to demonstrate the temperature stability of the device. During the experiments, the maximum injection flow rate is lowered to 10 µl/min with 2 µl/min increment and the time elapsed between different flow rate levels is set to be 60 seconds in order to minimize the amount of glycerol flow into the device from the syringe at room temperature, 23oC, which is not heated by the hotplate. At each temperature state, the device is first filled with sample and kept at least 20 minutes to assure the thermal equilibrium is attained. In addition, simultaneous viscosity measurement and optical imaging of the mixture of pyrogallol and NaOH aqueous solution is performed to demonstrate the real-time sample imaging capability of the device due to its optical transparency. In the experiment, flow rates of the sample fluid are controlled by the syringe pump with the same procedures of the water calibration. The entire device is placed on a stereomicroscope (M165FC, Leica Microsystems, Wetzlar, Germany) equipped with a digital camera (EOS-500D, Canon, Japan) for image recording. Viscosity and color variation are measured at the 10th, 30th minute and once an hour until the sixth hour as the chemical reaction is proceeding. For demonstration of the device viscosity measurement capability for non-Newtonian fluids, a Xanthan gum solution exploited for the experiment. During the experiment, the flow rate is increased up to 10 µl/min from static with 1 µl/min increment. To better characterize the effect of shear thinning


Page 5 of 17 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


that mainly occurs in low shear rate region, the lower flow rates, comparing to those used for viscosity measurements of the Newtonian fluids, are employed in the experiments. Since the sample is more viscous at low shear rates, it takes more time to complete the deformation of the elastic membrane. Therefore, the first two steps with flow rates of 1 and 2 µl/min are maintained for 300 seconds, while the remaining steps are maintained for 150 seconds. The entire experiment is also conducted in an environment with constant temperature of 23oC. To further demonstrate the usage of the device for viscosity measurement of practical biological samples, a human blood sample is tested in the experiments. During the experiments, the flow rate is increased up to 9 µl/min from static with 1 µl/min increment and each flow rate level is maintained for 100 seconds. The viscosity measurements are first performed under room temperature at 23oC. In order to mimic physiological flow and temperature conditions in human body, measurements under 37oC are also performed. All the viscosity measurements are performed in three different devices for statistical analysis. For comparison, the viscosity is also measured using a commercially available stress-controlled Physica Rheometer MCR 501 (Anton-Paar, Graz, Austria) equipped with a titanium Concentric-Cylinder Measurement System (CC-MS) under 37oC with approximately 1 ml of sample volume. The measurement procedure is performed according to ISO-3219.

RESULTS AND DISCUSSION Device Calibration The device is first calibrated using deionized water as a fluid sample, and raw data of the pressure sensor output voltage in time domain under different flow rates is shown in Figure 2(a). The actual flow rates within the microfluidic channel are experimentally confirmed to be similar to the designated ones set in the syringe pump as shown in Figure S-1 in the Supporting Information. During the measurement, approximate 300 µl of sample is injected into the device, and the volume can be reduced by decreasing the flow rates and the number of steps. In addition, dead volume of the device and the connected tubing is approximate 151 µl and 98 µl, respectively. The dead volume of the device can be further minimized by shortening the length of the bottom microfluidic channel to fit the specific range of viscosity as derived in the Supporting Information and shown in Figure S-2. In addition, the dead volume of the tubing can be minimized by optimizing the fluidic interconnections. The transient voltage response of the device takes less than 10 seconds to reach 90% of its steady state value between each flow rate level. The shorter transient time can be achieved for desired applications by further optimizing the device design, such as increasing the rigidity of the membrane and reducing the diameter of the pressure transduction hole. The measured voltage shifts in the last ten seconds at each applied flow rate is averaged and plotted in Figure 2(b) to investigate the relation between the output voltage shift and the flow rate. The result shows great linearity (R2>0.999 for linear regression) between the voltage shift and the flow rate, which suggests constant flow resistance during the water injection into the microfluidic channel as derived in equation (2). According to equation (6), the constant flow resistance

implies the constant viscosity of the water sample and constant geometries of the microfluidic channel under various flow rates. The constant viscosity agrees with Newtonian nature of water, and the agreement indicates that the assumptions of small membrane deformation (i.e. linear plate theory32) and small resistance change in the Wheatstone bridge setup are valid. The result also suggests that the increased base-tocuring agent ratio of the PDMS mixture prevents the channel wall within the device from expanding by hydrostatic pressure37.From the linear regression analysis, the pressure sensitivity of water is 0.1762 mV‧min/µl that is used to estimate the viscosity of fluids. In addition, the long-term stability of the device is also confirmed by device calibration using pressurized nitrogen gas for more than a week as shown in Figure S-3 in the Supporting Information. Newtonian Fluid Measurement The glycerol solutions with different concentrations are exploited as Newtonian fluidic samples with different viscosities. The same measurement procedures as the aforementioned calibration process are performed on the glycerol samples and the experimental condition is elaborated in the previous section. With concentration of the glycerol solution increases from 50% to 85%, the time required to reach the steady state prolongs from about 20 seconds to 80 seconds. The transient effect mainly results from the deformation of the connection tubing under pressure induced by the sample flow. The measured output voltage shifts of glycerol at different concentrations in the last ten seconds of each flow rate level are averaged and plotted in Figure 3(a). The result shows a larger voltage shift, induced by a larger pressure at the same flow rate, is caused by samples with higher viscosities, which agrees with equation (4). The high linearity of the device (R2>0.99 for linear regression), when measuring the glycerol solution with higher concentration, shows its ability to be applied for the measurements in a wide viscosity range. In addition, Figure 3(b) plots the relation between the pressure sensitivity and the sample viscosity reported in previous studies. The high linearity (R2>0.999 for linear regression) among water and glycerol solutions with different concentrations shows that the pressure sensitivity can be utilized to estimate the viscosity with high accuracy. Comparing the sensitivities of the glycerol solutions with the water, the viscosities of the glycerol samples can then be estimated by equation (7). In order to validate the measurement results, the calculated viscosities are compared to those reported in a previous study38. Table S-1 in the Supporting Information summarizes the viscosities measured using the developed microfluidic device and reported in the previous study, and their differences. The results show that the differences are no larger than 5.2%, which indicate that fluidic viscosity can be accurately measured using the microfluidic device with a broad viscosity range for various applications. In addition, higher viscosity can be measured by simply designing the geometries of the microfluidic channel as described in the Supporting Information to lower the pressure built up at upstream which exceeds the linear range of the pressure sensor. Apart from concentration, viscosity of glycerol solutions is also affected by temperature. A series of viscosity measurements are performed using 60% glycerol solution, from 4oC to 70oC, to demonstrate the device is capable of reliably perform-



ing viscosity measurements under different temperatures as shown in Figure 3(c). Total 30 µl of glycerol solution is injected into the device, which is initially filled with glycerol solution, during the measurement. The difference between the measured results and the previously reported value is no more than 5.7%. The measured results agree well with the previous study. The measured results are compared to the previously reported value38 and shown in the Table S-2 in the Supporting Information. Furthermore, the temperature compensation design of the device can be validated through the measured results. Possible minor cause of error may arise from the temperature gradient toward ambient surrounding which leads to an overestimate of the temperature of the sample. The aforementioned causes of error can be minimized by placing the whole experimental setup into a space with constant temperature, for example a temperature-controlled oven, to obtain more accurate results. Simultaneous viscosity measurement and sample imaging To show the capability of real-time imaging during measurement, viscosity of the mixture of pyrogallol and NaOH aqueous solution is measured for six hours. The sample undergoes an oxygen absorption chemical reaction, and the color turns to dark purple from colorless during the experimental period. Due to the long-term stability of the device, the viscosity measurement is performed using the same device during the six-hour measurement. The color variation of the sample is quantified by analyzing average gray scale value of the recorded image at each time point. Figure 4(a) shows the optical images and Figure 4(b) shows the quantitative viscosity measurement and optical imaging analysis results. The optical images show that the color of the mixture turns dark purple from colorless during the experiment. In addition, the plot shows that the viscosity increases about 3.72% of its original value (0.995cP) while the average gray scale value of the images reduce around 51%. The demonstrated device capability enables the study of aqueous sample viscosity and simultaneous optical observation during sample dynamic variation (e.g. chemical reactions), which can greatly help researchers investigate more information and physical insights comparing to the existing methods. Non-Newtonian Fluid Measurement For non-Newtonian fluid, a 1000 ppm Xanthan solution is tested in the experiment. The sample with the total volume of approximately 394 µl is injected into the device. The transient time is less than 90 seconds and gradually decreases when flow rate increases. The measured voltage shifts in the last ten seconds at each applied flow rate is averaged and plotted in Figure 5(a). According to equation (13), the power-law index, n, can be obtained through the regression line of a log-log plot of Figure 5(a) as shown in Figure 5(b). The estimated powerlaw index of the non-Newtonian sample is 0.51±0.03 that is comparable to the previous studies15,39. The highly linear relation in the log-log plot shown in Figure 5(b) indicates that the Xanthan gum solution is well characterized by the power-law model within the flow rate range and can therefore be able to give accurate results as described in previous studies15. The apparent viscosity values at various shear rates can then be estimated from the measured voltage shift in Figure 5(a) by equation (14) which is shown in Figure 5(c).

Page 6 of 17

Human Whole Blood Measurements under Different Temperatures A human whole blood sample is tested in the experiments to demonstrate usage of the device for practical biological applications. According to the aforementioned steps, the measured voltage is averaged and plotted in Figure 6(a) and the viscosity of blood can be estimated. Approximately 330 µl of blood sample is required for a complete experiment. The averaged time required to reach 90% of the steady state voltage value takes less than 50 seconds. Therefore, the required sample volume can be further reduced by decreasing the injection time if necessary. Viscosity of blood under 23oC and 37oC are shown in Figure 6(b). In order to validate the results measured under room temperature, the empirical relation between viscosity and temperature can be estimated that a 1oC increase in temperature results in a 2% decrease in blood viscosity40. Accordingly, the averaged percentage of increment in viscosity per degree Celsius is about 2.03% for the blood experiment which closely meets the empirical relation. The power-law index of the blood at 23oC and 37oC are approximately 0.681±0.018 (n=3) and 0.693±0.015 (n=3), respectively, which is expressed as mean±SD. Previous studies use the power-law model to characterize twenty blood samples collected from donors of the same age and gender. The experiments show that the power-law index is in the range of 0.716 to 0.66741, which agrees well with the result measured by the developed microfluidic device within the shear rate range from 1 to 90 1/s. In addition, comparison between the result measured by the device and the commercially available rheometer is shown in Figure 6(c). The results also show good agreement between the two measurements, which suggest great measurement accuracy achieved using the device. Therefore, the microfluidic viscometer is capable of performing accurate measurements on real biological samples under different temperatures with small sample volume, simple instrumentation and operation for practical biomedical applications. Furthermore, the optical transparency of the entire device provides simultaneous real-time imaging capability during the viscosity measurement as shown in the inset of Figure 6(c). A video in the electronic supplementary information shows a blood sample flowing in the microfluidic viscometer at different flow rates during the viscosity measurements. The demonstrated imaging capability is highly desired to study rheological properties for various fluidic samples, including biological ones. For example, hematocrit level usually varies (Hematocrit: 30 to 35%) during surgical operation, which leads to variation of blood viscosity. Therefore, the device can be applied to realtime monitor the blood viscosity value of a subject during surgery and even observe the blood condition in real time42.

CONCLUSION In this paper, we have successfully developed a fully disposable and optically transparent microfluidic viscometer based on ionic liquid electrofluidic circuit pressure sensor. The electrical circuit constructed using ionic liquid has desirable advantages, including: temperature stability, and long-term stability. Furthermore, the electrofluidic circuit sensing component can be directly fabricated within the device provides several advantages, such as effective cost (estimated to be less than USD$5/device), simple fabrication process, full disposa-


Page 7 of 17 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


bility and capability for large scale integration. In addition, the optical transparency of the device facilitates the simultaneous real-time monitoring of fluidic rheological behaviors. The accuracy and precision of the viscosity characterization with capability of measuring the Newtonian and non-Newtonian fluids under different temperatures are validated by testing a variety of solutions and biological samples. The results agree well with previous studies or a commercially available rheometer. Consequently, the developed microfluidic viscometer not only provides a simple and cost-effective instrument for viscosity characterization with small sample volumes, but also can be exploited in various applications where long-term observations and different temperature operations are required.

AUTHOR INFORMATION

15 16 17 18 19 20 21 22 23 24

Corresponding Author * Tel.: +886-2-2787-3138. Fax: +886-2-2787-3122. E-mail: [email protected] (Y.-C. Tung)

25 26 27

ACKNOWLEDGMENT We are grateful to National Central University Biophysics Core Facility for their support of this work. The protocol for the blood measurement experiment has been approved by the Institutional Review Boards of Tri-service General Hospital (2-103-05-092) and Academia Sinica (AS-IRB01-14014). This paper is based on work supported by the National Health Research Institutes (NHRI) in Taiwan under Innovative Research Grant (IRG) (EX106-10523EI), the Taiwan Ministry of Science and Technology (MOST 104-2221-E-001-015-MY3, 105-2221-E-001-002MY2) and the Academia Sinica Career Development Award.

28 29 30

31

32 33

REFERENCES 1 2 3

4 5 6

7 8 9

10

11 12 13 14

Rector, K. D.; Jiang, J.; Berg, M. A.; Fayer, M. D. J. Phys. Chem. B 2001, 105, 1081-1092. Ansari, A.; Jones, C. M.; Henry, E. R.; Hofrichter, J.; Eaton, W. A. Science 1992, 256, 1796-1798. Beece, D.; Eisenstein, L.; Frauenfelder, H.; Good, D.; Marden, M. C.; Reinisch, L.; Reynolds, A. H.; Sorensen, L. B.; Yue, K. T. Biochemistry 1980, 19, 5147-5157. Rosenson, R. S.; Mccormick, A.; Uretz,E. F. Clin. Chem. 1996, 42(8), 1189-1195. Letcher, R. L.; Chien, S.; Pickering, T. G.; Sealey, J.; Laragh, J. H. The American Journal of Medicine 1981, 70, 1195-1202. Devereux, R. B.; Drayer, J. I. M.; Chien, S.; Pickering, T. G.; Letcher, R. L.; DeYoung, J. L.; Sealey, J. E.; Laragh, J. H. Am. J. Cardiol. 1984, 54, 592-595. Ott, E. O.; Lechner,H.; Aranibar, A. Stroke 1974, 5, 330-333. Grotta, J.; Ackerman, R.; Correia, J.; Fallick, G.; Chang, J. Stroke 1982, 13, 296-301. Rafnsson, S.; Deary, I. J.; Whiteman, M. C.; Rumley, A.; Lowe, G. D. O.; Fowkes, F. G. R. Age Ageing 2010, 39, 217222. Tamariz, L. J.; Young, J. H.; Pankow, J. S.; Yeh, H.-C.; Schmidt, M. I.; Astor, B.; Brancati, F. L. Am. J. Epidemiol.2008, 168, 1153-1160. Irace, C.; Carallo, C.; Scavelli, F.; De Franceschi, M. S.; Esposito, T.; Gnasso, A. Diabetes Care 2014, 37, 488-492. Rantonen, P. J. F.; Meurman, J. H.; Acta Odontol Scand 1998, 56, 210-214. Gomez, J. E.; Thurston, G.B.; Biorheology 1993, 30(5-6), 409427. Roitman, E.V.; Dement'eva, I. I.; Kolpakov, R. E. Klin. Lab. Diagn 1995, 4, 29-31.

34 35 36 37 38 39 40 41 42

Srivastava, N.; M. A. Burns Anal. Chem. 2006, 78, 1690-1696. Han, Z.; Tang, X.; Zheng, B. J. Micromech. Microeng 2007, 17, 1828-1834. Guillot, P.; Panizza, P.; Salmon, J.-B.; Joanicot, M.; Colin, A.; Langmuir 2006, 22, 6438- 6445. Kang, K.; Lee, L. J.; Koelling, K. W. Exp. Fluids 2005, 38, 222–232. Pan, L.; Arratia, P. E.; Microfluid Nanofluidics 2012, 14(5), 885-894. Chevalier, J.; Ayela, F. Rev. Sci. Instrum. 2008, 79, 076102. Wu, C.-Y.; Liao, W.-H.; Tung, Y.-C. Lab Chip 2011, 11, 1740-1746. Wu, C.-Y.; Lu, J.-C.; Liu, M.-H.; Tung, Y.-C. Lab Chip 2012, 12, 3943-3951. Liu, M.-H.; Shih, H.-C.; Wu, J.-G; Weng, T.-W.; Wu, C.-Y.; Lu, J.-C.; Tung, Y.-C. Lab Chip 2013, 13, 1743. Lin, C.-H.; Wang, C.-K.; Chen, Y.-A.; Peng, C.-C.; Liao, W.H.; Tung, Y.-C. Lab Chip 2016, 6, 36425. Pandey, S. Anal. Chim. Acta 2006, 556, 38-45. Baker, G. A.; Baker, S. N.; Pandey, S.; Bright, F. V. Analyst 2005,130, 800-808. Liu, J.; Jonsson, J. A.; Jiang, G. TrAC, Trends Anal. Chem. 2005, 24(1), 20-27. Plechkova, N. V.; Seddon, K. R. Chem. Soc. Rev. 2008, 37, 123-150. Unger, M. A.; Chou, H. P.; Thorsen, T.; Scherer, A.; Quake, S. R.; Science, 2000, 288, 113–116. Samir I. Abu-Eishah (2011). Ionic Liquids Recycling for Reuse, Ionic Liquids - Classes and Properties, Prof. Scott Handy (Ed.), ISBN: 978-953-307-634-8, InTech. Mosadegh, B.; Kuo, C.-H.; Tung, Y.-C.; Torisawa, Y.; Bersano-Begey, T.; Tavana, H.; Takayama, S. Nat Phys 2010, 6, 433-437. Son, Y. Polymer 2007, 48, 632-637. Munson, B. R.; Young, D. F.; Okiishi, T. H.; Huebsch, W. W. Fundamentals of fluid mechanics, 6th ed.; Wiley: Hoboken, 2010. Hartnett, J. P.; Kostic, M.; Adv. Heat Transfer 1989, 19, 247356. Chen, Y.-A.; King, A.D.; Shih, H.-C.; Peng, C.-C., Wu, C.-Y.; Liao, W.-H.; Tung, Y.-C. Lab Chip 2011, 11, 3636. Katzbauer, B. Polym. Degrad. Stab. 1998, 59, 81-84. Gervais, T.; El-Ali, J.; Gunther, A.; Jensen, K. F. Lab Chip 2006, 6, 500–507. Segur, J. B.; Oberstar, H. E. Ind. Eng. Chem. Res. 1951, 43(9), 2117-2120. Speers, R. A.; Tung, M. A. J. Food Sci. 2006, 51(1), 96-98. Rand, P. W.; Lacombe, E.; Hunt, H. E.; Austin, W. H. J. Appl. Physiol. 1964, 19(1), 117-122. Elblbesy, M. A.; Hereba, A. T. Appl. Phys. Res. 2016, 8(2). Aronson, H. B.; Cotev, S.; Magora, F.; Borman, J. B.; Merin, G. Br. J. Anaesth. 1974, 46, 722.



Page 8 of 17

Figure 2. Device calibration using water. (a) Voltage in time domain under various flow rates during the calibration process. (b) Averaged voltage shift at different flow rate. Data are expressed as mean±sd. (n=3) Figure 1.Illustration of the microfluidic device. (a) The schematic diagram of the microfluidic device. (b) The working principle of the variable sensing resistor which converts mechanical pressure to electrical resistance variation. (c) The schematic diagram of the experimental setup.


Page 9 of 17 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 4. Imaging the sample as performing the viscosity measurement. (a) Optical images of the sample at different time points. (b) Viscosity of the sample and the corresponding analyzed gray scale value. The viscosity values are expressed as mean±sd. (n=3)

Figure 3. Viscosity measurement of Newtonian fluid. (a) Averaged voltage shift of glycerol solutions with different concentration at different flow rates. (b) Establishment of the relation between flow resistance and viscosity. (c) Experimental results under various temperatures. Data are expressed as mean±sd. (n=3)



Figure 5. 1000 ppm Xanthan gum viscosity measurement. (a) Averaged voltage shift at various flow rates of xanthan gum solution. (b) Log-log plot of the averaged voltage shift and the correspondingly applied flow rate. (c) The curve of measured viscosity versus different shear rate. Data are expressed as mean±sd. (n=3)

Page 10 of 17

Figure 6. Human whole blood viscosity measurements. (a) Averaged voltage shift at various flow rates of human whole blood. (b) Comparison between different temperatures. (c) Comparison of the experimental result to a commercially available rheometer. The inset shows a real-time image of human red blood cell inside the microfluidic channel. Data are expressed as mean±sd. (n=3).

Table 1. Geometric constants a*, b* and f* for fluidic channels with different aspect ratios.


Page 11 of 17 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1.Illustration of the microfluidic device. (a) The schematic diagram of the microfluidic device. (b) The working principle of the variable sensing resistor which converts mechanical pressure to electrical resistance variation. (c) The schematic diagram of the experimental setup. 84x124mm (200 x 200 DPI)



Figure 2. Device calibration using water. (a) Voltage in time do-main under various flow rates during the calibration process. (b) Averaged voltage shift at different flow rate. Data are expressed as mean±sd. (n=3) 82x108mm (200 x 200 DPI)


Page 12 of 17

Page 13 of 17 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3. Viscosity measurement of Newtonian fluid. (a) Aver-aged voltage shift of glycerol solutions with different concentra-tion at different flow rates. (b) Establishment of the relation be-tween flow resistance and viscosity. (c) Experimental results un-der various temperatures. Data are expressed as mean±sd. (n=3) 82x176mm (200 x 200 DPI)



Figure 4. Imaging the sample as performing the viscosity measure-ment. (a) Optical images of the sample at different time points. (b) Viscosity of the sample and the corresponding analyzed gray scale value. The viscosity values are expressed as mean±sd. (n=3) 82x108mm (200 x 200 DPI)


Page 14 of 17

Page 15 of 17 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 5. 1000 ppm Xanthan gum viscosity measurement. (a) Averaged voltage shift at various flow rates of xanthan gum solu-tion. (b) Log-log plot of the averaged voltage shift and the corre-spondingly applied flow rate. (c) The curve of measured viscosity versus different shear rate. Data are expressed as mean±sd. (n=3) 82x159mm (200 x 200 DPI)



Figure 6. Human whole blood viscosity measurements. (a) Averaged voltage shift at various flow rates of human whole blood. (b) Comparison between different temperatures. (c) Comparison of the experimental result to a commercially available rheometer. The inset shows a real-time image of human red blood cell inside the microfluidic channel. Data are expressed as mean±sd. (n=3). 82x155mm (200 x 200 DPI)


Page 16 of 17

Page 17 of 17


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Electrofluidic Circuit-Based Microfluidic Viscometer ... - ACS Publications

Recommend Documents