See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/259634315

Droplet formation behavior in a microfluidic device fabricated by hydrogel

molding

Article  in  Microfluidics and Nanofluidics · September 2014

DOI: 10.1007/s10404-013-1327-1

CITATIONS READS

8 368

5 authors, including:

Hirotada Hirama Hiroyuki Moriguchi

National Institute of Advanced Industrial Science and Technology The University of Tokyo
17 PUBLICATIONS   66 CITATIONS    68 PUBLICATIONS   728 CITATIONS   

SEE PROFILE SEE PROFILE

Toru Torii
The University of Tokyo
66 PUBLICATIONS   2,839 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Next-generation Agricultural System through the Use of IT View project

Encapsulation with a droplet View project

All content following this page was uploaded by Hirotada Hirama on 10 May 2015.

The user has requested enhancement of the downloaded file.

*Manuscript
Click here to download Manuscript: [revision 2]manuscript for submission.docx
Click here to view linked References

1
2
3 Droplet formation behavior in a microfluidic device fabricated by
4 hydrogel molding
5
6
7 Takahiro Odera, Hirotada Hirama*, Jo Kuroda, Hiroyuki Moriguchi, and Toru Torii
8
9
10
Department of Human and Engineered Environmental Studies, Graduate School of Frontier Sciences,
11
12 The University of Tokyo, 5-1-5, Kashiwanoha, Kashiwa-shi, Chiba 277-8563, Japan.
13
14 E-mail: hhirama@dt.k.u-tokyo.ac.jp
15 Tel/Fax: +81-4-7136-4654
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
61 1 / 21
62
63
64
65
 1
2
3 Abstract
4 We describe the behavior of droplet formation within 3D cross-junctions and 2D T-junctions with
5
6 various cross-sectional geometries that were manually fabricated using the hydrogel-molding
7 method. The method utilizes wire-shaped hydrogels as molds to construct 3D and 2D microchannel
8
9 structures. We investigated the flow patterns and droplet formation within the microchannels of these
10
microfluidic devices. Despite being fabricated manually, the microchannels with 3D cross-junctions
11
12 and 2D T-junctions were reproducible and formed highly monodispersed droplets. Additionally, the
13
14 sizes of the droplets formed within the microchannels could be predicted using an experimental
15 formula. This technique of droplet formation involves the use of a device fabricated by hydrogel
16
17 molding. This method is expected to facilitate studies on droplet microfluidics and promote the use
18
of droplet-based lab-on-a-chip technologies for various applications.
19
20
21
22 Keywords: Droplet microfluidics, 3D microchannel, Agarose gel, Lithography-free,
23 Polydimethylsiloxane
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
61 2 / 21
62
63
64
65
 1
2
3 1. Introduction
4 Droplet microfluidics has been studied for many decades (Kawakatsu et al. 1996), and numerous
5
6 droplet formation methods, including the use of T-junctions (Thorsen et al. 2001; Nisisako et al.
7 2002; Okushima et al. 2004), flow-focusing (Anna et al. 2003; Utada et al. 2005; Nie et al. 2005),
8
9 and electrostatic manipulation (Choi et al. 2007), have been developed for various applications, such
10
as particle synthesis (Nisisako et al. 2004; Nisisako et al. 2012; Zhang et al. 2012; Hirama et al.
11
12 2013; Aketagawa et al. 2013; Muluneh and Issadore 2013) and theoretical studies of droplet
13
14 formation in microchannels of various configurations (Kobayashi et al. 2004a; Samie et al. 2013).
15 Droplets form because of a balance between the interfacial tension and shear forces that are caused
16
17 by the branched microchannels in microfluidic devices and the drop in pressure (Garstecki et al.
18
2006). Because droplet formation in a microchannel is subject to the effects of the microchannel’s
19
20 cross-sectional shape, droplet formation using microchannels with different shapes has been
21
22 investigated in detail (Kobayashi et al. 2002; Kobayashi et al. 2004a, b). Polydimethylsiloxane
23 (PDMS) microfluidic devices have previously been fabricated to form droplets using
24
25 photolithography, which is a cost-effective and rapid prototyping technology (Sollier et al. 2011).
26 However, photolithography requires specialized facilities, such as a clean room with
27
28 photolithography equipment. A rapid, non-photolithographic prototyping method would contribute to
29
30 the advancement of microfluidics research in a variety of fields.
31 We previously reported a lithography-free, rapid prototyping method in which a hydrogel
32
33 mold was used to fabricate microchannels (Hirama et al. 2012). This method allows for the rapid
34 fabrication of flexible 2D and 3D microchannels without requiring photolithography because
35
36 hydrogel molds can be easily prepared in either glass capillaries or polytetrafluoroethylene (PTFE)
37
tubes. We previously presented this convenient fabrication method for the rapid prototyping of
38
39 microchannel devices. However, despite this convenience, the fabricated devices, specifically
40
41 microchannels with triangular cross-sections or 3D cross-junctions, have not been studied with
42 respect to droplet microfluidics. The fabrication of microchannels with a triangular cross-section is
43
44 crucial to facilitating droplet formation, and that of microchannels with 3D cross-junctions is crucial
45
to the mass production of Janus droplets or particles (Muluneh and Issadore 2013).
46
47 In this study, we investigated droplet formation behavior within several types of
48
49 microchannels (i.e., 2D channels, 3D channels, and channels with various cross-sectional
50 geometries) fabricated by the abovementioned method to expand the device’s applications, and we
51
52 report the effect of these architectures on the formation of microdroplets and experimental formulae
53 for predicting the sizes of the formed droplets.
54
55
56
57 2. Materials and Methods
58 2.1 2D microchannel fabrication
59
60
61 3 / 21
62
63
64
65
 1
2
3
The fabrication process, which is the same process used in our previously reported method involving
4 hydrogel molds (Hirama et al. 2012), is shown in Fig. 1. Briefly, an aqueous solution of 16% (v/v)
5
6 glycerol, 3% (w/v) agarose (Agarose L, Wako, Japan), and 0.05% (w/v) food dye (Red No. 102,
7 Tokyo Chemical Industry, Japan) was prepared and cured in a glass capillary to form a gel wire. The
8
9 diameter of the gel wire determined the size of the microchannel (Fig. 1 (a, b)). Gel wires were then
10
arranged into the desired channel design on pre-cured PDMS (10:1 weight ratio of base resin to
11
12 curing agent; SILPOT 184, Dow Corning, Japan). A 60-mm dish (60 mm×15 mm, Fisher Scientific,
13
14 USA) was used to contain the embedded PDMS. An additional layer of PDMS pre-polymer was cast
15 onto the pre-cured PDMS and agarose gel wires (Fig. 1 (c, d)), and the PDMS was cured at 55°C,
16
17 which is below the melting point of agarose, for 1 hour. After curing, a biopsy punch was used to
18
create holes in the PDMS device, and boiling water was flushed through the microchannels to
19
20 remove excess agarose gel (Fig. 1 (e, f)). Finally, the microchannels were treated with Sigmacoat
21
22 (Sigma, USA) to render them hydrophobic.
23 2.2 Fabrication of 3D microchannel cross-junctions
24
25 A 3D microchannel cross-junction was generated by crossing one gel wire over a second gel wire
26 (Fig. 1 (h) (top) and Fig. 2 (a)). Glass capillaries with inner diameters of 300 µm (special order,
27
28 Sun-yell Co., Japan) and 600 µm (G-1, Narishige, Japan) and polytetrafluoroethylene (PTFE) tubes
29
30 with inner diameters of 800 µm (Flon Ind. Co., Japan) were used to fabricate gel wires and 3D
31 microchannel cross-junction devices.
32
33 2.3 2D microchannel T-junctions with various cross-sectional geometries
34 The microchannel structure ultimately depends on the cross-sectional area of the glass capillary used
35
36 to prepare the gel wire. In this study, glass capillaries with circular, square, or triangular
37
cross-sections (special order, Takao Manufacturing Co., Japan) were used to prepare gel wires for
38
39 microchannels that would contain dispersed-phase solutions. Capillaries with a square
40
41 cross-sectional area were used to prepare microchannels containing continuous phase solutions (Fig.
42 2 (b)).
43
44 2.4 Microchannel droplet formation
45
2.4.1 3D microchannels
46
47 Mineral oil (Sigma, USA) supplemented with 1% (w/w) surfactant (SY glyster CRS-75; Sakamoto
48
49 Chemical Ind., Japan) with a viscosity of 21.5 mPa·s was used as the continuous phase. A solution of
50 1% (w/w) food coloring (Red 102 or Blue 1, Tokyo Chemical Industry, Japan) in either deionized
51
52 (DI) water or 2% (w/v) methylcellulose was used as the dispersed phase; the viscosities of the
53 solutions were 1.29 and 21.6 mPa·s, respectively. The interfacial tensions between the dispersed and
54
55 continuous phase (mineral oil) were 8.96 and 13.97 mN m-1 for the DI water and methylcellulose
56
57 solutions, respectively.
58 We used the 3D microchannel cross-junctions to detect the flow patterns of the dispersed
59
60
61 4 / 21
62
63
64
65
 1
2
3
and continuous phases and observe droplet formation. This geometry was achieved by varying the
4 diameter of the microchannel as well as the physical properties and flow rates of the fluid, as shown
5
6 in Table 1. The continuous phase flowed in the lower channel, and the dispersed phase flowed in the
7 upper channel (Fig. 2 (a)). Using DI water with food coloring as the dispersed phase, we measured
8
9 the biphasic droplet diameter with varying flow rates and channel diameters, as shown in Table 2.
10
The flow rate of the continuous phase was fixed, and the dispersed phase was varied at a constant
11
12 rate. For each condition, 50 droplets were measured.
13
14 2.4.2 2D microchannels
15 We measured the diameter of droplets formed in microchannel T-junctions with different
16
17 cross-sectional geometries by adjusting the flow rate of the dispersed phase to 0.4 ml h-1 (0.4-4.0 ml
18
h-1), 0.8 ml h-1 (0.4-4.0 ml h-1), and 2.0 ml h-1 (1.0-10.0 ml h-1). The values in parentheses represent
19
20 the flow rate of the continuous phase. DI water with food coloring was used as the dispersed phase.
21
22 The flow rate of the dispersed phase was fixed, and the flow rate of the continuous phase was varied
23 at a constant rate. For each condition, 50 droplets were analyzed.
24
25
26 3. Results and discussion
27
28 3.1 Microchannel fabrication
29
30 The diameters of the 3D cross-junctions were measured by filling the microchannels with colored
31 water and examining them under a microscope (Fig. 3 (a)). These diameters were determined to be
32
33 245 µm (9 µm) in 250-µm microchannels, 497 µm (10 µm) in 500-µm microchannels, and 743 µm
34 (3 µm) in 750-µm microchannels, with the standard deviation (SD) shown in parentheses. The
35
36 number of prepared and measured junctions was 5 each. The acute junction angle in the 2D
37
microchannel T-junctions was 87° (4.6°), with the SD shown in parentheses. The number of prepared
38
39 and measured junctions was 15 each. The coefficients of variation (CV) in both the diameter and the
40
41 angle were less than approximately 5%, which indicates that the hydrogel-based method yielded
42 reproducible microchannels. Table 3 lists the cross-sectional areas of microchannel T-junctions with
43
44 various geometries (measured as shown in Fig. 3(b)). The results confirm that the hydrogel-based
45
fabrication method generated uniform microchannels.
46
47 3.2 Flow patterns in 3D cross-junctions
48
49 We investigated specific flow patterns in 3D microchannel cross-junctions. In this study, the flow
50 pattern in 2D devices was not investigated because it has already been widely studied (Thorsen et al.
51
52 2001; Nisisako et al. 2002; Garstecki et al. 2006). The flow patterns and types of droplets formed are
53 shown in Fig. 5. Lee et al. (Lee et al. 2009) previously reported that microfluidic flow is controlled
54
55 by the balance of viscous stresses (described as viscosity) and capillary pressure (described as the
56
57 capillary number). In their study, the flow rate was expressed as the capillary number Ca =ηU/σ,
58 where η is the viscosity, σ is the interfacial tension, and U is the flow velocity.
59
60
61 5 / 21
62
63
64
65
 1
2
3
The biphasic flow and biphasic droplet formation of an immiscible two-layer flow were
4 stable, as shown in Fig. 5. However, alternating droplet generation, which denotes the formation of
5
6 two types of droplets (red and green colored droplets), resulted in unstable phenomena. The flow
7 conditions “unstable (biphasic flow)”, “unstable (biphasic droplet)”, “unstable (alternating droplet
8
9 generation)”, and “unstable” indicate flow patterns that transformed into another flow pattern within
10
5 minutes. Biphasic droplet generation in the 3D cross-junctions was observed for capillary numbers
11
12 within the following ranges: 2 × 10 -4 < CaC < 1 × 10-1 and 3 × 10-5 < Cad < 6 × 10-4, where CaC and
13
14 Cad are the capillary numbers of the continuous and dispersed phases, respectively. Therefore,
15 monodispersed biphasic droplets were expected to form in these capillary number ranges irrespective
16
17 of the microchannel diameter or the physical properties of the fluids.
18
Alternating droplet generation was only observed when the junction was filled with
19
20 mineral oil (continuous phase) prior to the addition of water (dispersed phase). The process in which
21
22 monodispersed droplets were alternatively generated for more than 5 minutes was classified as
23 monodispersed alternating droplet generation. The other instances were classified as unstable
24
25 alternating droplet generation. Unstable alternating droplet generation was observed in wider
26 channel devices (φ500 µm; conditions 2 and 3 in Table 1), and monodispersed generation was
27
28 observed only in wider channel devices using the dispersed phase with a lower viscosity (DI water;
29
30 condition 2 in Table 2). In addition, alternating droplet generation was observed within the range of
31 10-3 < Cac < 10-2. The droplet formation behavior in 3D microchannel devices has not been
32
33 previously studied, and thus, the data obtained in this study cannot be directly compared with other
34 data published in previous reports. However, we compared the obtained Ca range to previously
35
36 reported data for 2D alternating droplet generation by assuming the 3D device acted as a 2D
37
cross-junction device. The Ca range was observed to be within previously reported Ca ranges for
38
39 alternating droplet generation, and the behavior of alternating droplet generation with the devices
40
41 fabricated by hydrogel molding agreed with the results reported in previous studies (Zheng et al.
42 2004; Lee et al. 2010).
43
44 Figure 6 shows the relationship between the droplet diameter and the capillary number
45
ratio, which is defined as A= Ca d/Cac, where Cad and Cac are the capillary numbers of the dispersed
46
47 and continuous phases, respectively. In this study, the figures were created by using the
48
49 dimensionless diameter to describe the formed droplet size on a relative scale between the droplet
50 and microchannel dimensions. The dimensionless diameter, d nd, is equal to the mean diameter of the
51
52 droplets, dm, divided by the hydraulic diameter of the microchannel, defined as dh=4S/L, where S is
53 the cross-sectional area of the microchannel and L is the circumference of the channel. In this figure,
54
55 the dimensionless diameter of a droplet formed via 3D cross-junctions with channel diameter D
56
57 [µm] is represented as dnd=kD AlD, where kD and lD are constants. The correlation coefficients for the
58 three experimental conditions exceeded 0.9; therefore, the droplet diameter could be accurately
59
60
61 6 / 21
62
63
64
65
 1
2
3
estimated from this expression. The relationship between the flow rate of the dispersed phase and
4 droplet size in 3D microchannel devices has not been previously studied; therefore, the data obtained
5
6 in this study cannot be directly compared with other data from previous reports. However, we
7 compared the obtained relationship between the flow rate of the dispersed phase and droplet size to
8
9 previously reported data obtained from 2D cross-junction devices by assuming that the 3D devices
10
acted as 2D cross-junction devices based on the data shown in Fig. 6. In the 3D device, the droplet
11
12 sizes were proportional to the flow rates of the dispersed phase, and the proportionality constants
13
14 were 0.95-3.0. This result is very similar to previously published experimental results, which
15 indicate values of 1-2.52 (Xu et al. 2008).
16
17 3.3 Droplet formation in T-junctions with different cross-sectional geometries
18
Figure 7 shows the relationship between the droplet diameter and the ratio of capillary numbers,
19
20 A=Cad/Cac, in T-junctions with different cross-sectional geometries. In these 2D T-junctions, the
21
22 dimensionless diameter, dn, can also be accurately expressed as dnd= kD AlD. Additionally, the droplet
23 size was proportional to the flow rate of the dispersed phase in T-junction microchannels with a
24
25 square cross-section, and the proportionality constant was 1.7. This result is very similar to
26 previously published experimental and simulation results, which have indicated values of 1
27
28 (Garstecki et al. 2006) and 1.82 (Liu and Zhang 2009) for this constant. Therefore, the device
29
30 presented herein was confirmed to be a reliable rapid prototyping microfluidic device for various
31 applications.
32
33 The diameters of the droplets were similar in T-junctions with circular and square
34 cross-sections; however, these diameters were smaller than those of T-junctions with triangular
35
36 cross-sectional microchannels (Fig. 7). The variation in droplet size was effected by the difference in
37
channel depth; the circular and square cross-section depths were approximately 500 µm, and the
38
39 triangular cross-sectional depth was approximately 430 µm. According to a previous study (Chan et
40
41 al. 2005), the rapid increase in channel depth at a step of a microchannel causes a sudden decrease in
42 flow velocity, and a non-uniform Laplace pressure occurs at the corner of a junction with a step
43
44 structure, which then forms and pinches the “neck” of the dispersed phase and ultimately generates
45
droplets (Stone et al. 1986; Sugiura et al. 2001). These additional factors caused by the step structure
46
47 of the triangular cross-section gave rise to the difference in droplet size observed between the
48
49 circular and square cross-section devices and the triangular cross-section device. Although the angle
50 of the T-junction can also affect droplet formation, this effect was virtually negligible in this study.
51
52 This observation was made because according to a previously reported study, angles within a range
53 of 60-120° are not affected by the resulting droplet sizes (Yeom and Lee 2011), and the junction
54
55 angle in the device used in this study was approximately 90°.
56
57 The droplets formed in microfluidic devices fabricated by the hydrogel-molding method
58 were highly monodisperse with a CV <5%, as previously reported (Okushima et al. 2004). This
59
60
61 7 / 21
62
63
64
65
 1
2
3
fabrication method produced devices that were equally effective in forming uniform droplets
4 compared to existing microfluidic devices.
5
6
7 4. Conclusions
8
9
10 A hydrogel-molding method was used to fabricate 3D microchannel cross-junctions and T-junctions
11 with various cross-sectional geometries. The fabricated microchannels containing both 3D
12
13 cross-junctions and 2D T-junctions were reproducible and uniform, and they were used to evaluate
14
microchannel flow patterns. Monodisperse biphasic droplets could be generated when the capillary
15
16 numbers fell in the following ranges: 2 × 10-4 < CaC < 1 × 10-1 and 3 × 10-5 < Cad < 6 × 10-4. The
17
18 dimensionless diameter of the droplets formed in 3D cross-junctions and T-junctions increased
19 exponentially with the capillary number ratio of the continuous and dispersed phases. The diameters
20
21 of the droplets formed in microchannel T-junctions with various cross-sectional geometries were
22 compared. Microfluidic devices fabricated by hydrogel molding were used to develop a technique
23
24 for droplet formation. Our results are similar to those of previous studies; therefore, the devices
25
26 fabricated in this study can be used to prototype microfluidic devices reliably and rapidly. This
27 technique is expected to facilitate studies on droplet microfluidics and promote the use of
28
29 droplet-based lab-on-a-chip technologies for various applications.
30
31
32
33
34 References
35 Aketagawa K, Hirama H, Torii T (2013) Hyper-Miniaturisation of Monodisperse Janus
36
37 Hydrogel Beads with Magnetic Anisotropy based on Coagulation of Fe3O4
38 Nanoparticles. Journal of Materials Science and Chemical Engineering 1 (2):1-5.
39
40 doi:10.4236/msce.2013.12001
41
Anna SL, Bontoux N, Stone HA (2003) Formation of dispersions using “flow focusing” in
42
43 microchannels. Applied Physics Letters 82 (3):364. doi:10.1063/1.1537519
44
45 Chan EM, Alivisatos AP, Mathies RA (2005) High-temperature microfluidic synthesis of
46 CdSe nanocrystals in nanoliter droplets. Journal of the American Chemical Society
47
48 127 (40):13854-13861. doi:Doi 10.1021/Ja051381p
49
Choi WK, Lebrasseur E, Al-Haq MI, Tsuchiya H, Torii T, Yamazaki H, Shinohara E, Higuchi
50
51 T (2007) Nano-liter size droplet dispenser using electrostatic manipulation
52
53 technique. Sens Actuator A-Phys 136 (1):484-490. doi:DOI 10.1016/j.sna.2006.12.028
54 Garstecki P, Fuerstman MJ, Stone HA, Whitesides GM (2006) Formation of droplets and
55
56 bubbles in a microfluidic T-junction - scaling and mechanism of break-up. Lab on a
57 chip 6 (3):437-446. doi:Doi 10.1039/B510841a
58
59 Hirama H, Kambe T, Aketagawa K, Ota T, Moriguchi H, Torii T (2013) Hyper alginate gel
60
61 8 / 21
62
63
64
65
 1
2
3 microbead formation by molecular diffusion at the hydrogel/droplet interface.
4 Langmuir : the ACS journal of surfaces and colloids 29 (2):519-524.
5
6 doi:10.1021/la303827u
7 Hirama H, Odera T, Torii T, Moriguchi H (2012) A lithography-free procedure for fabricating
8
9 three-dimensional microchannels using hydrogel molds. Biomedical microdevices 14
10
(4):689-697. doi:10.1007/s10544-012-9649-4
11
12 Kawakatsu T, Kikuchi Y, Nakajima M (1996) Visualization of Microfiltration Phenomena
13
14 Using Microscope Video System and Silicon Microchannels. Journal of Chemical
15 Engineering of Japan 29 (2):399-401
16
17 Kobayashi I, Mukataka S, Nakajima M (2004a) CFD simulation and analysis of emulsion
18
droplet formation from straight-through microchannels. Langmuir : the ACS journal
19
20 of surfaces and colloids 20 (22):9868-9877. doi:Doi 10.1021/La0487489
21
22 Kobayashi I, Mukataka S, Nakajima M (2004b) Effect of slot aspect ratio on droplet
23 formation from silicon straight-through microchannels. Journal of colloid and
24
25 interface science 279 (1):277-280. doi:DOI 10.1016/j.jcis.2004.06.028
26 Kobayashi I, Nakajima M, Chun K, Kikuchi Y, Fukita H (2002) Silicon array of elongated
27
28 through-holes for monodisperse emulsion droplets. Aiche Journal 48 (8):1639-1644.
29
30 doi:DOI 10.1002/aic.690480807
31 Lee M, Park W, Chung C, Lim J, Kwon S, Ahn KH, Lee SJ, Char K (2010) Multilayer
32
33 deposition on patterned posts using alternating polyelectrolyte droplets in a
34 microfluidic device. Lab on a chip 10 (9):1160-1166. doi:Doi 10.1039/B919753b
35
36 Lee W, Walker LM, Anna SL (2009) Role of geometry and fluid properties in droplet and
37
thread formation processes in planar flow focusing. Physics of Fluids 21 (3).
38
39 doi:10.1063/1.3081407
40
41 Liu HH, Zhang YH (2009) Droplet formation in a T-shaped microfluidic junction. Journal of
42 Applied Physics 106 (3). doi:10.1063/1.3187831
43
44 Muluneh M, Issadore D (2013) Hybrid soft-lithography/laser machined microchips for the
45
parallel generation of droplets. Lab on a chip. doi:10.1039/c3lc50979f
46
47 Nie ZH, Xu SQ, Seo M, Lewis PC, Kumacheva E (2005) Polymer particles with various
48
49 shapes and morphologies produced in continuous microfluidic reactors. Journal of
50 the American Chemical Society 127 (22):8058-8063. doi:10.1021/ja042494w
51
52 Nisisako T, Ando T, Hatsuzawa T (2012) High-volume production of single and compound
53 emulsions in a microfluidic parallelization arrangement coupled with coaxial
54
55 annular world-to-chip interfaces. Lab on a chip 12 (18):3426-3435.
56
57 doi:10.1039/c2lc40245a
58 Nisisako T, Torii T, Higuchi T (2002) Droplet formation in a microchannel network. Lab on a
59
60
61 9 / 21
62
63
64
65
 1
2
3 chip 2 (1):24-26. doi:10.1039/b108740c
4 Nisisako T, Torii T, Higuchi T (2004) Novel microreactors for functional polymer beads.
5
6 Chemical Engineering Journal 101 (1-3):23-29. doi:10.1016/j.cej.2003.11.019
7 Okushima S, Nisisako T, Torii T, Higuchi T (2004) Controlled production of monodisperse
8
9 double emulsions by two-step droplet breakup in microfluidic devices. Langmuir :
10
the ACS journal of surfaces and colloids 20 (23):9905-9908. doi:10.1021/la0480336
11
12 Samie M, Salari A, Shafii MB (2013) Breakup of microdroplets in asymmetric T junctions.
13
14 Physical Review E 87 (5). doi:10.1103/PhysRevE.87.053003
15 Sollier E, Murray C, Maoddi P, Di Carlo D (2011) Rapid prototyping polymers for
16
17 microfluidic devices and high pressure injections. Lab on a chip 11 (22):3752-3765.
18
doi:10.1039/c1lc20514e
19
20 Stone HA, Bentley BJ, Leal LG (1986) An Experimental-Study of Transient Effects in the
21
22 Breakup of Viscous Drops. Journal of Fluid Mechanics 173:131-158. doi:Doi
23 10.1017/S0022112086001118
24
25 Sugiura S, Nakajima M, Iwamoto S, Seki M (2001) Interfacial tension driven monodispersed
26 droplet formation from microfabricated channel array. Langmuir : the ACS journal
27
28 of surfaces and colloids 17 (18):5562-5566. doi:Doi 10.1021/La010342y
29
30 Thorsen T, Roberts RW, Arnold FH, Quake SR (2001) Dynamic pattern formation in a
31 vesicle-generating microfluidic device. Physical Review Letters 86 (18):4163-4166
32
33 Utada AS, Lorenceau E, Link DR, Kaplan PD, Stone HA, Weitz DA (2005) Monodisperse
34 double emulsions generated from a microcapillary device. Science 308
35
36 (5721):537-541. doi:10.1126/science.1109164
37
Xu JH, Li SW, Tan J, Luo GS (2008) Correlations of droplet formation in T-junction
38
39 microfluidic devices: from squeezing to dripping. Microfluidics and Nanofluidics 5
40
41 (6):711-717. doi:10.1007/s10404-008-0306-4
42 Yeom S, Lee SY (2011) Dependence of micro-drop generation performance on dispenser
43
44 geometry. Exp Therm Fluid Sci 35 (8):1565-1574. doi:DOI
45
10.1016/j.expthermflusci.2011.07.008
46
47 Zhang J, Coulston RJ, Jones ST, Geng J, Scherman OA, Abell C (2012) One-Step Fabrication
48
49 of Supramolecular Microcapsules from Microfluidic Droplets. Science 335
50 (6069):690-694. doi:DOI 10.1126/science.1215416
51
52 Zheng B, Tice JD, Ismagilov RF (2004) Formation of droplets of in microfluidic channels
53 alternating composition and applications to indexing of concentrations in
54
55 droplet-based assays. Analytical chemistry 76 (17):4977-4982.
56
57 doi:10.1021/ac0495743
58
59
60
61 10 / 21
62
63
64
65
 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
61 11 / 21
62
63
64
65
 1
2
3
Figures
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 Fig. 1 Fabrication process of a PDMS microfluidic device using the hydrogel-molding method. (a)
32
33 Agarose gels are inserted into glass capillaries. (b) Agarose gels are cured and ejected from the
34 capillary. (c) Gel wires and gel chips are arranged on pre-cured PDMS sheets according to the
35
36 desired channel design. (d) Pre-cured PDMS is poured into the dish. (e) PDMS is cured by heating.
37
(f) Holes are punched into gel chips. (g) Excess agarose gel is washed away. (h) Microchannel
38
39 configuration. (i) Design of inlet and outlets for fluid injection and ejection.
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61 12 / 21
62
63
64
65
 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 Fig. 2 Schematic drawing of microchannel configuration. (a) A 3D microchannel cross-junction as
27
28 viewed from above (left) and along its cross-section (right). (b) A microchannel T-junction with
29
30 different cross-sectional geometries. In each device, one microchannel has a square cross-sectional
31 area and the other microchannel has either a circular, square, or triangular cross-sectional area.
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
61 13 / 21
62
63
64
65
 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 Fig. 3 Dimensions of fabricated devices. (a) Diameter of fabricated 3D cross-junctions. For the 3D
48
49 microchannels, the junction was approximated by a circle. (b) The cross-sectional geometry of
50 microchannel T-junctions. Uniform cross-sections were observed at each position. The scale bars are
51
52 500 µm.
53
54
55
56
57
58
59
60
61 14 / 21
62
63
64
65
 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 Fig. 4 Microchannel T-junctions with different cross-sectional geometries. (a) Photograph of gel
50 wires with square and triangular cross-sections. (b) Micrographs of T-junction cross-sections
51
52 obtained by phase contrast (top) and electron microscopy (bottom). The scale bars are 500 µm.
53
54
55
56
57
58
59
60
61 15 / 21
62
63
64
65
 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
Fig. 5 Flow pattern and droplet formation in 3D cross-junctions. The “biphasic flow” and “biphasic
39 droplets” consisted of a two-layer immiscible flow. “Alternating droplet generation” produced
40
41 alternating monodispersed droplets. These flow patterns persisted for more than 5 minutes. By
42 contrast, the flow conditions “unstable (biphasic flow)”, “unstable (biphasic droplet)”, “unstable
43
44 (alternating droplet generation)”, and “unstable” were flow patterns that transformed into another
45
flow pattern within 5 minutes. In the “unstable (transient)” flow condition, biphasic flow and
46
47 biphasic droplet formation were unstable and alternated. Circles indicate measurement points.
48
49
50
51
52
53
54
55
56
57
58
59
60
61 16 / 21
62
63
64
65
 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 Fig. 6 Relationship between the dimensionless droplet diameter and capillary number ratio of 3D
26 microchannel cross-junctions. The hydraulic diameters of the 250-, 500-, and 750-µm microchannels
27
28 were 250, 500, and 750 µm, respectively. Qd shows the flow rates of dispersed phase (mL/h). The
29
30 coefficient of variation (CV) in the droplet diameters at each point was less than 5%. The graph
31 shows log-log plots. The fits were obtained from the experimental results.
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
61 17 / 21
62
63
64
65
 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 Fig. 7 Relationship between the dimensionless droplet diameter and capillary number ratio in
50 microchannel T-junctions. Each graph shows the results from microchannel devices with different
51
52 cross-sectional geometries for the dispersed phase. The hydraulic diameters of the circular, square,
53 and triangular cross-section microchannel were 500, 500, and 289 µm, respectively. Qd shows the
54
55 flow rates of dispersed phase. The graphs show lin-log plots. The fits were determined from
56
57 experimental results.
58
59
60
61 18 / 21
62
63
64
65
 1
2
3 Table 1 Experimental conditions used to investigate flow patterns with 3D microchannel cross-junctions
4
5 Diameter of microchannel Continuous phase Dispersed phase (viscosity [mPa･s])
6
Condition 1 φ250 µm Mineral oil DI water (1.29)
7
8 Condition 2 φ500 µm Mineral oil DI water (1.29)
9
Condition 3 φ500 µm Mineral oil Methylcellulose (aq.) (21.6)
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
61 19 / 21
62
63
64
65
 1
2
3 Table 2 Experimental conditions used to measure the diameters of biphasic droplets
4
5 Diameter of the microchannel Flow rate of the continuous phase [ml/h] (interval) Flow rate of the dispersed phase [ml/h]
6 φ250 µm 0.1-1.0 (0.1) 0.2
7
8 0.1-1.0 (0.1) 0.4
9
1.0-10.0 (1.0) 1.0
10
11 φ500 µm 0.4-4.0 (0.4) 0.2
12
13 0.4-4.0 (0.4) 0.4
14 φ750 µm 1.0-10.0 (1.0) 1.0
15
16 2.0-20.0 (2.0) 2.0
17
2.0-20.0 (2.0) 4.0
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
61 20 / 21
62
63
64
65
 1
2
3 Table 3 Cross-sectional areas of microchannel T-junctions with
different geometries (corresponding to Fig. 3(b))
4
5
Cross-sectional area [mm2]
6
7 X [mm] Circle Square Triangle
8
9 0.0 0.175 0.241 0.121
10 7.5 0.172 0.246 0.119
11
12 15.0 0.168 0.244 0.119
13
22.5 0.172 0.242 0.114
14
15 30.0 0.173 0.244 0.120
16
17 Average 0.172 0.244 0.119
18 Standard deviation 0.003 0.002 0.003
19
20 CV [%] 1.48 0.76 2.27
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
61 21 / 21
62
63
64
65
View publication stats