Mixing Performance Evaluation for Commercially Available

Subscriber access provided by UB + Fachbibliothek Chemie | (FU-Bibliothekssystem)

Full Paper

Mixing Performance Evaluation for Commercially Available Micromixers using Villermaux-Dushman Reaction Scheme with IEM Model Joseph M Reckamp, Ashira Bindels, Sophie Duffield, Yangmu Chloe Liu, Eric Bradford, Eric M. Ricci, Flavien Susanne, and Andrew Rutter Org. Process Res. Dev., Just Accepted Manuscript • Publication Date (Web): 09 May 2017 Downloaded from http://pubs.acs.org on May 9, 2017

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.

Organic Process Research & Development 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 18

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

Organic Process Research & Development

Mixing Performance Evaluation for Commercially Available Micromixers using Villermaux-Dushman Reaction Scheme with IEM Model Joseph M. Reckampa*, Ashira Bindelsb,c, Sophie Duffieldc, Yangmu Chloe Liud, Eric Bradfordb, Eric Riccid, Flavien Susannec, and Andrew Rutterc a

c

Evonik Industries, 1650 Lilly Rd, Lafayette, IN 47909

b

lmperial College of London, SW7 2AZ, U.K.

GlaxoSmithKline, Product and Process Engineering, Stevenage, U.K. dGlaxoSmithKline, Product and Process Engineering, 709 Swedeland Road, King of Prussia, PA 19406, United States

*Corresponding author. [email protected]; 1650 Lilly Rd, Lafayette, IN 47909

ACS Paragon Plus Environment


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

For Table of Contents Only


Page 2 of 18

Page 3 of 18

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


Abstract The development of low-volume continuous processes for the pharmaceutical industry requires a greater understanding of mixing in microreactors. In this paper, numerous commercially available micromixers are evaluated using the Villermaux-Dushman reaction scheme and the interaction by exchange with the mean (IEM) mixing model to quantify the mixing time. The work presents the mixing times as a function of flow rate and energy dissipation for mixers including T-Mixers, Ehrfeld Mikrotechnik BTS micromixers, and Syrris Asia microchips. Highlights:

•

Villermaux-Dushman reaction scheme with IEM model used to calculate mixing times

•

Mixing times and energy dissipation of Ehrfeld, Syrris, and T-mixers presented

•

Significant differences observed in mixing efficiency between micromixers

Keywords: Mixing, Micromixers, Villermaux-Dushman, IEM

1. Introduction Batch processes dominate the pharmaceutical industry due to the strict regulatory environment and desire for rapid market introduction1-5. However, batch processes are often difficult to scale3. Regulators are now encouraging the development of continuous processing for its advantages in homogeneity, scale-up, and safety while also providing a solution to pricing pressures by reducing capital, labor, and inventory costs1-3,6-8. While continuous manufacturing is commonplace for high-volume regulated industries such as food and beverage, the low volumes of typical active pharmaceutical ingredients have generated a relatively novel market for micro-scale continuous reactor systems6,8-11. Miniaturization of



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

reactors, driven by key equipment manufacturers such as Ehrfeld Mikrotechnik BTS and Syrris Ltd, provides significant quality advantages in terms of enhanced heat transfer, mass transfer, and mixing efficiency2,6,8,9,11. Continuous processes often involve rapid reactions that are often limited by the rate of mixing, making mixing a critical design parameter, especially in chemical reactions where poor mixing would result in reduced selectivity and yield, or scenarios where “pockets” of high concentration reagents could results in precipitates and clogs that are detrimental to a flow reactor. Despite being a fundamental unit operation, mixing issues are often difficult to identify and isolate due to their reliance on knowledge of chemical reaction kinetics. Therefore, numerous publications have developed methods to compare mixer efficiency on generic systems including the use of dyes, pH changes, and competing reactions6,7,12. Due to potential bias in dye measurements and the requirement for transparent mixers for quantification, competitive reactions have become standard for mixing analysis as they lead to a quantitative method of characterizing mixer performance6-7. The development of competitive reaction systems for mixing characterization has been dominated by Bourne and Villermaux. Common reaction systems include the Diazo Coupling, Bourne III, Bourne IV, and Villermaux-Dushman reaction systems6,10,13-18. The Villermaux-Dushman system has generally been accepted as the standard for micromixer characterization due to its advantages in terms of in-line analysis by UV spectrophotometry, quantification of mixing times via reaction kinetics, inexpensive reagents, and the ability to adjust the analytical range through the reagent concentrations6,10,19-20. Quantitative analysis of the Villermaux-Dushman system relies on the selection of a mixing model describing the mechanistic properties of the integration of the two fluid streams. Numerous publications describe and compare the common mixing models, including the interaction by exchange with the mean (IEM), incorporation, and engulfment models21-22. Since the debate continues as to the accuracy of the


Page 4 of 18

Page 5 of 18

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


various phenomenological models and the models exhibit relatively low variation of calculated mixing times, the simple IEM model was chosen for this paper to provide a baseline quantifiable for comparison of mixing times. The IEM model assumes that the two streams exchange mass with a characteristic time constant equal to the mixing time. This paper uses the Villermaux-Dushman reaction system coupled with the IEM quantification model to characterize the mixing performance of a wide range of commercially available micromixers. The micromixers analyzed comprise common vendors such as Ehrfeld Mikrotechnik and Syrris, employing common mixing principles that can be separated into four categories: (1) simple contacting; (2) flow obstacles; (3) split and recombine; (4) multilamination. Table 1. Commercially available micromixers investigated in this paper. Images used with permission granted by Ehrfeld Mikrotechnik BTS, Lonza, Syrris Ltd, and Swagelok Company. © Ehrfeld Mikrotechnik BTS 2016, © Lonza 2016, © Syrris Ltd 2016, © Swagelok Company 2016.

Micromixer Supplier

Mixing Principle Mixer Information Split and Recombine Internal Volume: 103µL (06) and 173µL (15) Channel width: 0.6mm (06) and 1.5 mm (15)

Cascade 06, 15 0216-3 Ehrfeld Mikrotechnik BTS

Lonza TG Large 1701-2380 Ehrfeld Mikrotechnik BTS

Flow Obstacles Internal Volume: 1.2mL Mixer nominal width : 0.5 mm

Lonza Multiinjection SZ 1701-1642 Ehrfeld Mikrotechnik BTS

Flow Obstacles Internal Volume: 400µL Nominal width :200-600 µm

Slit Plate LH 2 0113-4 Ehrfeld Mikrotechnik BTS

Multilamination Internal Volume: 40µL Narrowest diameter: 25-100 µm depending on the mixing and aperture plate



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

Split and Recombine Internal Volume: 6.25µL Mixing channel dimension: 50 x 125 µm

Micromixer 2101411 Syrris Asia

Flow Obstacles Internal Volume: 62.5µL and 250µL Mixing channel dimension: 85 x 220 µm (62.5 µL); 250 x 300 µm (250 uL) Simple Contacting 1/8” T-Mixer 1/16” T-Mixer Narrowest diameter : 2.3 mm (SS-200-3) ; 1.3 mm (SS-100-3)

62.5µL and 250µL Microreactors 2100141, 2100143 Syrris Asia

T-Mixer SS-200-3, SS-100-3 Swagelok

2. Experimental 2.1 Experimental Setup Mixing characterization with the Villermaux-Dushman requires precise and consistent flow for enhanced accuracy of results. The setup consisted of two calibrated syringe pumps (Teledyne Isco 260D or Syrris Asia Syringe Pump depending on flow rates) that delivered the reagent streams to the mixer. Pressure sensors (Keller EV-120) were attached to Ehrfeld Mikrotechnik cells at the inlet and outlet of the micromixer to evaluate energy dissipation in the mixing zone. 1/16” PFA tubing connected the micromixer to the pumps and an optical flow cell for UV-Vis spectrometer measurement with a minimum of 1m of tubing between the micromixer outlet and the UV-Vis measurement to ensure reaction completion and equilibration. The tubing length required was confirmed with a second UV-Vis spectrometer located after additional residence time displaying an equivalent triiodide concentration. An additional 1m (2m total length) of tubing length was provided for equilibration when using the Tpiece mixers as the mixing was not complete within the T-piece. Optical fibers linked the flow cell to a UV-Vis spectrometer (Zeiss).


Page 6 of 18

Page 7 of 18

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. Experimental setup process flow diagram for Villermaux-Dushman reaction scheme.

Prior to the experiments, the UV-Vis was calibrated by preparing solutions of known triiodide concentrations, measuring the absorbance values at 353nm wavelength with a 625nm baseline correction to correct for any shift in the baseline measurement. Before each run, a reference spectra of water was obtained. The experimental procedure consisted of setting the two pumps to their desired equivalent flow rates and waiting for steady-state as identified by inlet and outlet pressure sensors (Keller EV-120). As steady-state was approached, UV-Vis absorbance data was collected for 90 seconds to allow for statistical confirmation of the results. The procedure was repeated for each flow rate and micromixer combination with duplicates and inter-run cleaning performed to confirm lack of iodine adsorbance to the flow cell. The Beer-Lambert law relating the measured absorbance and triiodide concentration only displays linearity over a low absorbance range10. Thus, the reagent concentrations must be adjusted to maintain readings within the valid linear range. A range of concentration sets proposed by Commenge and Falk shown in Table 2 provided the basis for this experiment. These reagent concentration sets afford distinct ranges of potential mixing times with negligible variance of the mixing time observed in changing concentration set.



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

Page 8 of 18

Table 2. Concentration sets (1, 1b, 1c, 2, 2b, and 2c) of various reagent concentrations (mol/L) to achieve different mixing 6

ranges developed by Commenge and Falk for the Villermaux-Dushman reaction system .

Concentration (mol/L) [HClO4] [KI] [KIO3] [NaOH] [H3BO3]

1

1b

1c

2

2b

2c

0.03 0.032 0.006 0.09 0.09

0.06 0.032 0.006 0.09 0.09

0.04 0.032 0.006 0.09 0.09

0.015 0.016 0.003 0.045 0.045

0.03 0.016 0.003 0.045 0.045

0.02 0.016 0.003 0.045 0.045

2.2 Mixing Characterization The Villermaux-Dushman reaction system consists of a series of competitive reactions including (1) a nearly instantaneous neutralization reaction, (2) a fast iodide-iodate reaction, and (3) an iodine-iodide reaction to form triiodide ions detectable by the UV-Vis spectrometer6,10:

+ →

5 + + 6 → 3 + 3

+

(1)

(2)

(3)

Where the first reaction was assumed to be instantaneous (k1=1011L/mol·s)23, k2 is a function of the ionic strength µ of the reaction mixture defined as log10(k2)=9.28105-3.664√μ when µ 0.166 M. Additionally, log10(Keq)=

+

7.355 −

2.575 !"# $, with T in Kelvin and Keq in L/mol. 6,10 In the case of ideal mixing, reaction (1) would proceed to consume all of the acid present instantaneously, minimizing the slower reaction (2) conversion, which results in an absence of triiodide ions. Poor mixing results in local concentrations of the acid, in which case the instantaneous reaction (1) with borate ions does not completely consume the acid, allowing reaction (2) to proceed to yield iodine,


Page 9 of 18

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


which is detected by UV-Vis as the triiodide ion. Therefore, the mixing time can be calculated by the extent of reaction (2) as observed through the equilibrium relationship in reaction (3) to form the triiodide ion. Kinetic rate expressions and the equilibrium expression for the three reactions have been characterized extensively in literature, with the first reaction forward reaction rate vastly exceeding the reverse dissociation rate and the second reaction being a function of the ionic strength of the solution6,10,23. UV-Vis absorbance data combined with the Beer-Lambert Law provides a calculation of the extent of each reaction as the reactions are assumed to have completed and equilibrated prior to measurement. The final concentrations of each component can be analyzed by the IEM model to calculate the amount of time required for complete mixing (tm). The IEM model assumes perfect plug flow behavior with no back-mixing present to ensure interaction of only equivalent age streams, which is supported by estimated Bodenstein numbers that ranges from 900 to 2x1010. Therefore, each species (i) can be described by two differential equations (subscript 1 for acid-rich stream; subscript 2 for buffer-rich stream) relating the concentrations to their mean concentrations () and reaction rates (R)6,10: %&',

%)

%&', %)

=

+&' ,&',

=

+&' ,&',

)-

)-

+ ./,"

(4)

+ ./,

(5)

Where the mean concentration can be calculated as a function of concentration of each stream and the acid flow rate volume fraction (α)10: < 1/ > = 31/," + (1 − 3)1/,


(6)


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

The mixing time calculation via the Villermaux-Dushman reaction system and IEM model integrates the mixing progress from introduction of the two streams until complete reagent consumption by using Equations 4 and 5 for each system component to best fit the UV-Vis spectrometer result. 3. Results and Discussion The mixing time results from the Villermaux-Dushman reaction system with IEM model are displayed in Table 3, which is color coded based on the magnitude of the mixing time. As a comparison of mixing time and residence time (Table 4) confirms, the mixing times are orders of magnitude shorter than residence time in all the mixers excluding T-pieces, supporting the assumption that the reactions are complete prior to UV measurement. The results from this study are for equivalent (1:1) acid and buffer stream flow rates; however, additional work investigating the mixing times for non-equivalent flow rates is ongoing. It is clear that each flow rate range requires different mixers. While the Syrris Asia microchips provide good mixing at low flow rates, their flow rate range is limited due to pressure ratings. At a greater flow rate range, Ehrfeld mixers provide a distinct advantage, achieving mixing times less than a millisecond.


Page 10 of 18

Page 11 of 18

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


Table 3 Mixing time results tm (ms) for commercially available mixers based on the Villermaux-Dushman reaction system evaluated with the IEM model. Reproducibility of mixing time results shown with Cascade 06 mixer. *This study was performed at equivalent 1:1 flow rate ratio between acid and buffer streams.

Total Flow Rate (mL/min)* 0.1 Ehrfeld Cascade 06

0.2

0.5

1

2

4

8

12

16

20

30

65 ± 5

41 ± 5

37 ± 4 177

19 ± 3 133

7.8 ± 1.4 87

4.5 ± 1.5 81

3.4 ± 1.0 50

2.4 ± 0.6 42

1.8 ± 0.2 33

0.7 ± 0.1 19

94

35

19

7.4

5.5

3.1

Ehrfeld Cascade 15 Ehrfeld Lonza TG Large Ehrfeld Lonza Multiinjection SZ

82

25/25µm mixing; 25µm aperture Ehrfeld Slit Plate

52

40

11

3.6

1.1

0.7

117

98

91

5.1

0.9

0.6

0.5

0.3

0.03

114

70

26

2.5

1.0

0.6

0.4

0.2

30

4.6

1.2

0.9

1.4

1.3

0.6

121

98

29

6.9

1.9

1.2

0.8

0.4

10

4.9

3.4

25

12

15

25/25µm mixing; 50µm aperture 100/25µm mixing; 25µm aperture 100/25µm mixing; 50µm aperture Syrris Asia Micromixer

45

38

25

Syrris Asia 62.5µL Microreactor

33

22

5.4

Syrris Asia 250µL Microreactor

45

33

25

1/8" TPiece

1/16" tubing

565

448

211

128

128

160

96

34

1/8" tubing

970

896

766

744

645

670

319

102

1/16" TPiece Key (ms) >700 400700 100500 10-100

1/16" tubing

1220

1100

645

216

232

201

191

164

1/8" tubing

1910

1300

941

88

59

43

33

24

1.0-10

Mixing Performance Evaluation for Commercially Available

Recommend Documents