Crystallization Kinetics of Cephalexin Monohydrate in the Presence of

Jul 22, 2019 - Martha A. Grover ... Publication History ... The Supporting Information is available free of charge on the ACS ... details of mixing fa...
1 downloads 0 Views 1MB Size
Subscriber access provided by Nottingham Trent University

Article

Crystallization kinetics of cephalexin monohydrate in the presence of cephalexin precursors Matthew A. McDonald, Grant D Marshall, Andreas S Bommarius, Martha A. Grover, and Ronald W. Rousseau Cryst. Growth Des., Just Accepted Manuscript • DOI: 10.1021/acs.cgd.9b00429 • Publication Date (Web): 22 Jul 2019 Downloaded from pubs.acs.org on July 23, 2019

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 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 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.

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 49 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

Crystal Growth & Design

Crystallization kinetics of cephalexin monohydrate in the presence of cephalexin precursors Matthew A. McDonald, Grant D. Marshall, Andreas S. Bommarius, Martha A. Grover, Ronald W. Rousseau*

*School of Chemical and Biomolecular Engineering, Georgia Institute of Technology, Atlanta GA, 30332

The solubility and crystallization kinetics of cephalexin monohydrate, an important β-lactam antibiotic, are reported for the first time for an aqueous solution. The nucleation and growth kinetics are determined with online process analytical technologies (PAT) and offline image analysis to support implementation of a continuous process. Effects of two co-solutes, 7-aminodesacetoxycephalosporanic acid (7-ADCA) and phenylglycine methyl ester (PGME), on nucleation and growth are examined. These co-solutes are precursors in the enzymatic synthesis of cephalexin.

ACS Paragon Plus Environment

1

Crystal Growth & Design 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

Primary nucleation of cephalexin was unaffected by the co-solutes, but 7-ADCA, even in small quantities ( 1 Well Mixed

0

0

200

400

600

Stir rate,

(rpm)

ACS Paragon Plus Environment

28

800

1000

Page 29 of 49 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

Crystal Growth & Design

Figure 9. A mixing diagram for the crystallizer geometry used to determine cephalexin crystallization kinetics. Experiments marked with an open circle (○) were well mixed and were used to fit the kinetic model. Experiments marked with a cross (×) showed poor mixing and experiments marked with a plus (+) showed significant foaming.

In Figure 9, the boundary between the poorly mixed (orange) and well mixed (yellow) regions was linearly regressed with the cross (×) points; the boundary compares favorably with the theoretical boundary based on transition to turbulent flow as determined by the measured (solid red line) and extrapolated (dashed red line) Reynolds Number, Re, defined in equation 15.

 D2 Re  

(15)

At suspension densities as low as 1.5%, crystallization in a quiescent vessel led to a “sherbet-like” slurry that would not flow under force of gravity. Similar non-flowability is observed in the non-mixed regions in low-rpm experiments. The non-flowability can be explained through the lens of yield stress; below a critical mixing intensity the slurry behaves as a solid, however above a certain stress the material yields and flows as a

ACS Paragon Plus Environment

29

Crystal Growth & Design 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

slurry. Formation of dead zones is a function of both rpm and solids concentration. Dead zones, where the stress due to mixing is below the yield stress, may be expected when the Reynolds number indicates predominately laminar flow (Re < 3,000);38 however, the red line in Figure 8 shows the observed sherbet behavior occurred at Reynolds numbers greater than 3,000. The density used in equation 15 was calculated with the known solids fraction, using a measured liquid density of 1.005 g/mL for a saturated solution at pH 4.5 and a solid density of 1.403 g/mL. The viscosity was measured with a capillary viscometer, which was not practical for solids fraction greater than 1%. Measuring the viscosity of suspensions is difficult; the measured viscosity is an estimate useful for qualitative interpretation of the mixing results. A linear relation between solids fraction and viscosity was measured (supplemental Figure S8) and used to construct the solid portion of the red line in Figure 9 and then extrapolated to higher solids fractions (dashed red line). Dead zones were accompanied by a decrease in the observed rate of desupersaturation, and experiments showing poor mixing behavior were excluded from the parameter fitting routine.

ACS Paragon Plus Environment

30

Page 30 of 49

Page 31 of 49 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

Crystal Growth & Design

The boundary of the foaming (blue) region is the convex hull of the experiments that generated foam, with the zero suspension density point being determined with the Froude number, Fr. The foam generated at high stirrer speeds consisted of air bubbles and entrained mother liquor stabilized by the needle-shaped crystals.

Engineered

needle-shaped particles have been used to stabilize surfactant foams,39 and the presence of the crystals in the cephalexin system is critical to the foam stability. Foaming was only observed once crystallization began, and foam accumulated and did not collapse when mixing ceased; FBRM results showed a reduction in counts of nearly 30% which qualitatively indicated a significant amount of solids was sequestered in the foam (supplemental Figure S9). The Froude number indicates formation of aerating surface vortices when Fr is of order unity; in Figure 9, Fr = 1 is indicated at ω = 930, however the real transition is less well-defined than the abrupt change at ω = 930. Fr, defined in Equation 16 (where gr is the acceleration due to gravity), predicted significant aeration at greater than 900 rpm, however Figure 9 shows significant foaming occurred at stir rates as low as 650 rpm (blue region). Foam forming experiments were also

ACS Paragon Plus Environment

31

Crystal Growth & Design 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 32 of 49

excluded in the parameter fitting routine, leaving a narrow window in which to operate (yellow region in Figure 9).

Fr  

D 2 gr

(16)

Equation 7 assumes a power law relationship between stirring energy input ε and secondary nucleation rate, B, hence the need to examine a range of stirring rpm. Within the narrow range of well-mixed ω there was no obvious correlation between ε (as calculated with Equation 14) and B. Therefore, the stir power input exponent, p, was set to zero to avoid overfitting the limited data. All further experiments were conducted at 400 rpm for consistency. With p = 0 and no co-solute present, the secondary nucleation rate and growth rate were simultaneously fit (see the next section for discussion of growth rate). The secondary nucleation prefactor, kB, had a value 2.98 × 105 min-1 L-1 (95% C.I: 1.01 × 105 4.95 × 105). The secondary nucleation exponent, b, had a fit value of 0.99, which was not statistically different from b = 1. Since there is a theoretical justification for b to be unity,36 b was set to 1 and a confidence interval was not calculated. The suspension

ACS Paragon Plus Environment

32

Page 33 of 49 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

Crystal Growth & Design

density exponent, m, was found to be 0.46 (0.30 0.62). A value of m < 1 suggests secondary nuclei form from crystal-impeller collisions, but not all crystals breed nuclei upon impeller impact; perhaps only crystals of a certain size breed secondary nuclei in such events. Since secondary nucleation is defined in terms of the growth rate (see Equation 7), the effect of co-solutes on secondary nucleation was mathematically captured by the effect of co-solutes on growth rate, which are discussed in the next section. That the growth rate equation completely captured the co-solute dependence may indicate that secondary nucleation occurs by attrition of larger crystals,40 however these data are too sparse to exclude other, less pronounced, effects of co-solutes on secondary nucleation, e.g. surface nucleation.41 In any case, the functional dependence of secondary nucleation and growth on supersaturation is empirical and appears to be the same. In Figure 10 (left), the CSD is shown for pure cephalexin, with PGME (30 mM) as co-solute, and with 7-ADCA (30 mM) as co-solute. A Kolmogorov-Smirnov test (which is agnostic to the underlying distribution) failed to reject the null hypothesis that the

ACS Paragon Plus Environment

33

Crystal Growth & Design 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

measured CSD for pure cephalexin and measured CSD for PGME co-solute sampled different underlying distributions at the 95% confidence level. However, as is apparent from the yellow bars at long crystal lengths in Figure 10 (left), a Kolmogorov-Smirnov test rejected the null hypothesis (at the 95% confidence level) that the pure CSD and 7ADCA co-solute CSD came from the same distribution. This result was repeated in several data sets. When fitting the experiments with 7-ADCA co-solute, the value of m (the suspension density exponent in Equation 7) was consistently at the high end of the confidence interval determined for pure cephalexin monohydrate (0.30 0.62). Simulated experiments in silico showed that larger values of m result in wider spread CSDs. It is possible that the presence of 7-ADCA increased the sensitivity of secondary nucleation to the suspension density (i.e. increased m), however, there was no statistically significant difference between the values of m determined from fitting pure experiments and co-solute containing experiments.

Therefore, it was concluded that co-solutes did

not have a substantial impact on the kinetics of secondary nucleation (beyond the proportionality between B and G established in Equation 7), especially when compared to the impact of co-solutes on the kinetics of crystal growth.

ACS Paragon Plus Environment

34

Page 34 of 49

Page 35 of 49

Cephalexin supersaturation

0.4

Normalized frequency

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

Crystal Growth & Design

No co-solute 30 mM PGME 30 mM 7-ADCA

0.3 0.2 0.1 0

0

20

40

60

80

100

120

No co-solute 30 mM PGME 30 mM 7-ADCA

2.50

2.00

1.50

1.00

0

Crystal Length ( m)

20

40

60

80

Time (minutes)

Figure 10. Left, the final CSD of cephalexin monohydrate without co-solute (blue), with 30 mM PGME co-solute (red), and 30 mM 7-ADCA co-solute (yellow). Right, the desupersaturation of cephalexin without co-solute (blue circle), with 30 mM PGME cosolute (red diamond), and 30 mM 7-ADCA co-solute (yellow square).

Growth kinetics The different co-solutes had different impacts on crystal growth. Having determined that the co-solute impact on nucleation was negligible compared to the impact on growth (these two phenomena were determined simultaneously while fitting crystallization experiments with the PBM), the decrease in desupersaturation rate observed in the presence of co-solutes, particularly 7-ADCA, was the result of growth inhibition by the co-solutes. Two models of growth inhibition were considered, a

ACS Paragon Plus Environment

35

Crystal Growth & Design 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 36 of 49

competitive model based on co-solute and solute competing for surface adsorption sites,42, 43 and a step-pinning model based on adsorbed co-solutes impeding advancing growth steps.44 Both models captured the decreased rate of desupersaturation, however only the step-pinning model accounts for the observed incomplete desupersaturation seen in the presence of co-solutes. In Figure 10 (right) the cephalexin supersaturation stopped decreasing at long times in the presence of PGME (red diamonds) or 7-ADCA (yellow squares), a behavior captured by the step pinning model of growth.

kG  S  Scr  S  1 g 1 S  Scr G 0 S  Scr 

(17)

In the step-pinning mechanism of crystal growth inhibition, a modified growth rate, defined in Equation 17, is used when the supersaturation is above a critical supersaturation, Scr, and zero growth rate is used when the supersaturation is below

Scr.

The

critical

supersaturation

prevents

the

solution

from

completely

desupersaturating. Figure 11 shows how the critical supersaturation increased as the concentration of each co-solute increased.

ACS Paragon Plus Environment

36

Page 37 of 49

1.4

Critical supersaturation, Scr

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

Crystal Growth & Design

7-ADCA

1.3

PGME

1.2 1.1 1 0

20

40

60

80

100

120

Co-solute concentration (mM) Figure 11. The critical supersaturation, Scr, as a function of co-solute concentration. The lines are best fits of Equation 18. The error bars are 95% confidence intervals.

The Langmuir adsorption isotherm was used to fit the dependence of the critical supersaturation on co-solute concentration because the step-pinning mechanism is based on co-solutes adsorbing onto the growing surface. The critical supersaturation depends on the energy required to create new cephalexin monohydrate surface, γ, the area of the growth unit, a, the average spacing between adsorption sites, l, the Langmuir equilibrium coefficient, K, and the concentration of co-solute, ccs according to Equation 18. Using the value of γ calculated from primary nucleation and the fit of Scr in

ACS Paragon Plus Environment

37

Crystal Growth & Design 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 38 of 49

Figure 11, the ratio a l evaluated to approximately 1 (fit value of 1.15) for 7-ADCA and 0.5 (fit value of 0.61) for PGME, suggesting that 7-ADCA adsorption sites and cephalexin monohydrate growth units are equally spaced but PGME adsorption sites are spaced over every two growth units. 7-ADCA, containing both an acid and an amine group may be able to form hydrogen bonds with the cephalexin monohydrate crystal in two locations, while PGME, containing only an amine group, may only form a hydrogen bond with the growing crystal in one location. However this interpretation is not definite as there is substantial uncertainty surrounding the values of a l , especially for PGME where the fit to Equation 18 is poor (see Figure 11). PGME and 7-ADCA may also preferentially adsorb onto different crystal faces.

Scr 

 aKccs

kTl 1  Kccs 

1

(18)

The needle-like morphology causes a small crystal surface area (the faces on the ends) to account for the majority of desupersaturation. At high supersaturation, the strong adsorption of 7-ADCA onto the fast growing faces suggests the possibility that 7ADCA would be incorporated into the cephalexin monohydrate crystal; PGME on the

ACS Paragon Plus Environment

38

Page 39 of 49 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

Crystal Growth & Design

other hand would not be expected to incorporate into the crystal to the same extent. PXRD did not show any peak shifts in the pattern for cephalexin crystallized with the 7ADCA co-solute, indicating the 7-ADCA and cephalexin are not crystallizing together (as a co-crystal) in a distinct crystal lattice. HPLC showed that crystals produced with 7ADCA co-solute contained 7-ADCA in proportion with the amount of initial 7-ADCA, even after washing with cold water (see supplemental Figure S10). However the amount of 7-ADCA was always less than 0.2% by weight of the final product after washing, indicating washed crystals were of pharmaceutical quality. PGME was not detected in washed crystals produced with PGME co-solute. Discussion of continuous manufacturing All aspects of the design of an end-to-end CM process for cephalexin need to consider the crystallization kinetics of cephalexin monohydrate. The rate of primary nucleation is slow, even at considerable supersaturation, therefore the process should rely on secondary nucleation, either by recycling product crystals or using a well-mixed crystallizer. Secondary nucleation did not depend on the mixing intensity within the range of mixing intensities that were feasible for the studied crystallizer geometry.

ACS Paragon Plus Environment

39

Crystal Growth & Design 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

However, the stir rate was important because cephalexin monohydrate suspensions have a propensity to form poorly mixed “sherbets” with insufficient agitation and difficultto-handle foams with excess agitation. In the enzymatic synthesis of cephalexin, one of the parameters than can be used to optimize the process is the reactant concentration ratio of PGME and 7-ADCA.7 This work suggests that, from the viewpoint of crystallization, it is best to run the reaction lean in 7-ADCA. Complete consumption of 7-ADCA in the reaction will lead to lower crystal growth inhibition, rendering more of the cephalexin recoverable (by crystallization) in a shorter time. A 7-ADCA lean process is further supported by the finding that 7-ADCA is more likely than PGME to incorporate into the cephalexin monohydrate crystal, albeit at part per thousand concentrations, and that 7-ADCA is more valuable than PGME. Conclusion The crystallization kinetics of cephalexin monohydrate are reported for the first time in an aqueous environment, summarized in Table 2, and the effect of co-solutes that would be found in the enzymatic synthesis of cephalexin were investigated. The

ACS Paragon Plus Environment

40

Page 40 of 49

Page 41 of 49 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

Crystal Growth & Design

solubility was determined over a range of pH values (4 to 8) and at two temperatures (5.0 °C and 25.0 °C), crystallization kinetics were explored only at 25.0 °C. Primary nucleation of cephalexin monohydrate occurred slowly, even at high supersaturation (S > 2), and was not impacted by the presence of PGME or 7-ADCA co-solutes. The secondary nucleation of cephalexin birthed many more crystals than primary nucleation (see Supplemental Figure S11) and depended on the growth rate and suspension density, but did not appear to depend on the mixing power in the regime accessible in this crystallizer geometry. At low mixing intensity the needle-shaped crystals were prone to forming dead zones while at high mixing intensities the needle-shaped crystals stabilized the formation of difficult-to-handle foam. Table 2. Kinetic parameters for the crystallization of cephalexin monohydrate with 95% confidence interval. Confidence interval values marked “ns” indicate parameters that were not significantly different from a theoretically justified value, therefore they have been fixed at the theoretically predicted value. Parameter name

symbol [units]

Primary nucleation prefactor kJ [min-1 L-1]

Value

Lower bound Upper bound

2.54

1.39

ACS Paragon Plus Environment

41

9.00

Crystal Growth & Design 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

Primary nucleation

Page 42 of 49

Bo

1.79

1.28

2.88

kB [min-1 L-1]

2.98 ×

1.01 × 105

4.95 × 105

exponential Secondary nucleation prefactor

105

b

1.0

ns

ns

m

0.46

0.30

0.62

Mixing power exponent

p

0

ns

ns

Growth prefactor

kG [μm min-1]

6.52

5.16

7.43

growth exponent

g

2.0

ns

ns

Secondary nucleation exponent Suspension density exponent

The crystal growth rate was inhibited by both co-solutes, though 7-ADCA was a more potent inhibitor. 7-ADCA inhibited cephalexin crystal growth by adsorbing onto the growing faces and impeding the advancement of growth steps. These results in aggregate all suggest that an end-to-end CM process for cephalexin would be most efficient in a 7-ADCA-lean configuration. Acknowledgements The authors thank Dr. Giovanni Maria Maggioni for his discussions on nucleation kinetics. We thank the Specialty Separations Center at Georgia Institute of Technology for supporting this project. We gratefully acknowledge funding and insight provided by

ACS Paragon Plus Environment

42

Page 43 of 49 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

Crystal Growth & Design

the US FDA under Grant U01FD006484-01. M.A. McDonald also acknowledges support from the NSF I/UCRC Center for Pharmaceutical Development (CPD) (Grant 1540017).

ACS Paragon Plus Environment

43

Crystal Growth & Design 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

References

[1] Van Boeckel, T. P., Gandra, S., Ashok, A., Caudron, Q., Grenfell, B. T., Levin, S. A., and Laxminarayan, R. (2014) Global antibiotic consumption 2000 to 2010: an analysis of national pharmaceutical sales data, The Lancet Infectious Diseases

14, 742-750. [2] Laxminarayan, R. (2014) Antibiotic effectiveness: Balancing conservation against innovation, Science 345, 1299-1301. [3] Adamo, A., Beingessner, R. L., Behnam, M., Chen, J., Jamison, T. F., Jensen, K. F., Monbaliu, J.-C. M., Myerson, A. S., Revalor, E. M., and Snead, D. R. (2016) Ondemand continuous-flow production of pharmaceuticals in a compact, reconfigurable system, Science 352, 61-67. [4] Mascia, S., Heider, P. L., Zhang, H., Lakerveld, R., Benyahia, B., Barton, P. I., Braatz, R. D., Cooney, C. L., Evans, J., and Jamison, T. F. (2013) End‐to‐End Continuous Manufacturing of Pharmaceuticals: Integrated Synthesis, Purification, and Final Dosage Formation, Angewandte Chemie International Edition 52, 12359-12363. [5] Cole, K. P., Groh, J. M., Johnson, M. D., Burcham, C. L., Campbell, B. M., Diseroad, W. D., Heller, M. R., Howell, J. R., Kallman, N. J., Koenig, T. M., May, S. A., Miller, R. D., Mitchell, D., Myers, D. P., Myers, S. S., Phillips, J. L., Polster, C. S., White, T. D., Cashman, J., Hurley, D., Moylan, R., Sheehan, P., Spencer, R. D., Desmond, K., Desmond, P., and Gowran, O. (2017) Kilogram-scale prexasertib monolactate monohydrate synthesis under continuous-flow CGMP conditions,

Science 356, 1144. [6] Jolliffe, H. G., and Gerogiorgis, D. I. (2016) Plantwide design and economic evaluation of two Continuous Pharmaceutical Manufacturing (CPM) cases: Ibuprofen and artemisinin, Computers & Chemical Engineering 91, 269-288. [7] McDonald, M. A., Bommarius, A. S., Rousseau, R. W., and Grover, M. A. (2019) Continuous reactive crystallization of β-lactam antibiotics catalyzed by penicillin

ACS Paragon Plus Environment

44

Page 44 of 49

Page 45 of 49 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

Crystal Growth & Design

G acylase. Part I: Model development, Computers & Chemical Engineering 123, 331-343. [8] Encarnación-Gómez, L. G., Bommarius, A. S., and Rousseau, R. W. (2016) Crystallization Kinetics of Ampicillin Using Online Monitoring Tools and Robust Parameter Estimation, Industrial & Engineering Chemistry Research 55, 21532162. [9] Ottens, M., Lebreton, B., Zomerdijk, M., Rijkers, M., Bruinsma, O. S. L., and Van der Wielen, L. A. M. (2001) Crystallization kinetics of ampicillin, Industrial &

Engineering Chemistry Research 40, 4821-4827. [10] Kubota, N., Yokota, M., and Mullin, J. (1997) Supersaturation dependence of crystal growth in solutions in the presence of impurity, Journal of Crystal Growth

182, 86-94. [11] Kubota, N. (2001) Effect of impurities on the growth kinetics of crystals, Crystal

Research and Technology 36, 749-769. [12] Fevotte, F., and Fevotte, G. (2010) A method of characteristics for solving population balance equations (PBE) describing the adsorption of impurities during crystallization processes, Chemical Engineering Science 65, 3191-3198. [13] Youshko, M. I., Chilov, G. G., Shcherbakova, T. A., and Švedas, V. K. (2002) Quantitative characterization of the nucleophile reactivity in penicillin acylasecatalyzed acyl transfer reactions, Biochimica et Biophysica Acta (BBA)-Proteins

and Proteomics 1599, 134-140. [14] McDonald, M. A., Bommarius, A. S., and Rousseau, R. W. (2017) Enzymatic reactive crystallization for improving ampicillin synthesis, Chemical Engineering

Science 165, 81-88. [15] Griffin, D. J., Grover, M. A., Kawajiri, Y., and Rousseau, R. W. (2014) Robust multicomponent IR-to-concentration model regression, Chemical Engineering

Science 116, 77-90. [16] Griffin, D. J., Grover, M. A., Kawajiri, Y., and Rousseau, R. W. (2015) Combining ATR-FTIR and FBRM for feedback on crystal size, In 2015 American Control

Conference (ACC), pp 4308-4313, IEEE.

ACS Paragon Plus Environment

45

Crystal Growth & Design 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

[17] Li, J., Trout, B. L., and Myerson, A. S. (2015) Multistage Continuous MixedSuspension, Mixed-Product Removal (MSMPR) Crystallization with Solids Recycle, Organic Process Research & Development 20, 510-516. [18] Lai, T.-T. C., Cornevin, J., Ferguson, S., Li, N., Trout, B. L., and Myerson, A. S. (2015) Control of Polymorphism in Continuous Crystallization via Mixed Suspension Mixed Product Removal Systems Cascade Design, Crystal Growth &

Design 15, 3374-3382. [19] Li, H., Kawajiri, Y., Grover, M. A., and Rousseau, R. W. (2017) Modeling of nucleation and growth kinetics for unseeded batch cooling crystallization,

Industrial & Engineering Chemistry Research 56, 4060-4073. [20] Abu Bakar, M. R., Nagy, Z. K., and Rielly, C. D. (2010) Investigation of the Effect of Temperature Cycling on Surface Features of Sulfathiazole Crystals during Seeded Batch Cooling Crystallization, Crystal Growth & Design 10, 3892-3900. [21] Li, H., Grover, M. A., Kawajiri, Y., and Rousseau, R. W. (2013) Development of an Empirical Method Relating Crystal Size Distributions and FBRM Measurements,

Chemical Engineering Science 89, 142-151. [22] Li, H., Kawajiri, Y., Grover, M. A., and Rousseau, R. W. (2014) Application of an Empirical FBRM Model to Estimate Crystal Size Distributions in Batch Crystallization, Crystal Growth & Design 14, 607-616. [23] Maggioni, G. M., and Mazzotti, M. (2015) Modelling the stochastic behaviour of primary nucleation, Faraday Discussions 179, 359-382. [24] Maggioni, G. M., Bosetti, L., dos Santos, E., and Mazzotti, M. (2017) Statistical Analysis of Series of Detection Time Measurements for the Estimation of Nucleation Rates, Crystal Growth & Design 17, 5488-5498. [25] Burton, W.-K., Cabrera, N., and Frank, F. (1951) The growth of crystals and the equilibrium structure of their surfaces, Philosophical Transactions of the Royal

Society of London A: Mathematical, Physical and Engineering Sciences 243, 299-358. [26] Barthe, S. C., Grover, M. A., and Rousseau, R. W. (2008) Observation of polymorphic change through analysis of FBRM data: Transformation of paracetamol from form II to form I, Crystal Growth & Design 8, 3316-3322.

ACS Paragon Plus Environment

46

Page 46 of 49

Page 47 of 49 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

Crystal Growth & Design

[27] Hutton, K., Mitchell, N., and Frawley, P. J. (2012) Particle size distribution reconstruction: The moment surface method, Powder Technology 222, 8-14. [28] Wold, S., Sjöström, M., and Eriksson, L. (2001) PLS-Regression: a Basic Tool of Chemometrics, Chemometrics and Intelligent Laboratory Systems 58, 109-130. [29] Lange, L., Lehmkemper, K., and Sadowski, G. (2016) Predicting the Aqueous Solubility of Pharmaceutical Cocrystals As a Function of pH and Temperature,

Crystal Growth & Design 16, 2726-2740. [30] Bada, J. L. (1969) The pKa of a weak acid as a function of temperature and ionic strength: An experiment using a pH meter, Journal of Chemical Education 46, 689. [31] Xiao, Y., Tang, S. K., Hao, H., Davey, R. J., and Vetter, T. (2017) Quantifying the Inherent Uncertainty Associated with Nucleation Rates Estimated from Induction Time Data Measured in Small Volumes, Crystal Growth & Design 17, 2852-2863. [32] Kennedy, A. R., Okoth, M. O., Sheen, D. B., Sherwood, J. N., Teat, S. J., and Vrcelj, R. M. (2003) Cephalexin: a channel hydrate, Acta Crystallographica

Section C: Crystal Structure Communications 59, o650-o652. [33] Davison, A. C., and Hinkley, D. V. (1997) Bootstrap methods and their application, Vol. 1, Cambridge university press. [34] Lacmann, R., Herden, A., and Mayer, C. (1999) Kinetics of nucleation and crystal growth, Chemical Engineering & Technology: Industrial Chemistry‐Plant

Equipment‐Process Engineering‐Biotechnology 22, 279-289. [35] Armenante, P. M., Mazzarotta, B., and Chang, G.-M. (1999) Power consumption in stirred tanks provided with multiple pitched-blade turbines, Industrial &

Engineering Chemistry Research 38, 2809-2816. [36] Ayazi Shamlou, P., Jones, A. G., and Djamarani, K. (1990) Hydrodynamics of secondary nucleation in suspension crystallization, Chemical Engineering

Science 45, 1405-1416. [37] Szilágyi, B., Agachi, P. Ş., and Lakatos, B. G. (2015) Numerical analysis of crystallization of high aspect ratio crystals with breakage, Powder Technology

283, 152-162.

ACS Paragon Plus Environment

47

Crystal Growth & Design 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

[38] Sinnott, R. K. (1993) CHAPTER 10 - Equipment Selection, Specification and Design, In Coulson and Richardson's Chemical Engineering (Second Edition) (Sinnott, R. K., Ed.), pp 352-438, Pergamon, Amsterdam. [39] Alargova, R. G., Warhadpande, D. S., Paunov, V. N., and Velev, O. D. (2004) Foam superstabilization by polymer microrods, Langmuir 20, 10371-10374. [40] Garside, J., Rusli, I., and Larson, M. A. (1979) Origin and size distribution of secondary nuclei, AIChE Journal 25, 57-64. [41] Tai, C. Y., Wu, J.-F., and Rousseau, R. W. (1992) Interfacial supersaturation, secondary nucleation, and crystal growth, Journal of Crystal Growth 116, 294306. [42] Kubota, N., and Mullin, J. (1995) A kinetic model for crystal growth from aqueous solution in the presence of impurity, Journal of Crystal Growth 152, 203-208. [43] Albon, N., and Dunning, W. (1962) Growth of sucrose crystals: determination of edge energy from the effect of added impurity on rate of step advance, Acta

Crystallographica 15, 474-476. [44] Parker, R. L. (1970) Crystal Growth Mechanisms: Energetics, Kinetics, and Transport, In Solid State Physics (Ehrenreich, H., Seitz, F., and Turnbull, D., Eds.), pp 151-299, Academic Press.

ACS Paragon Plus Environment

48

Page 48 of 49

Page 49 of 49

For Table of Contents Use Only Crystallization kinetics of cephalexin monohydrate in the presence of cephalexin precursors Matthew McDonald, Grant Marshall, Andreas Bommarius, Martha Grover, Ronald

1.4 7-ADCA

1.3 1.2 PGME

1.1 1 0

50

100

Co-solute concentration (mM)

Normalized frequency

Rousseau Critical supersaturation, S cr

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

Crystal Growth & Design

Pure PGME 7-ADCA

0.04 0.03 0.02 0.01 0

0

50

100

Crystal Length ( m)

150

The co-solutes 7-ADCA and PGME inhibit growth and secondary nucleation of cephalexin crystals below a critical supersaturation, Scr. 7-ADCA is a more potent inhibitor than PGME. Image analysis used to construct size distributions show 7-ADCA resulting in larger average crystal size while PGME does not affect crystal size.

The supporting material document contains information regarding: in-situ IR spectroscopy, simulated crystallization observed with FBRM, offline image analysis, bootstrapping for construction of confidence intervals, details of mixing failure modes, purity analysis, and details of nucleation.

ACS Paragon Plus Environment

49