Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Nucleation in continuous flow cooling sonocrystallization for coiled capillary crystallizers

Nucleation in continuous flow cooling sonocrystallization for coiled capillary crystallizers Nucleation in continuously operated capillary coiled cooling crystallizers is experimentally investigated under the influence of ultrasound. It was found that there is no sharp boundary but rather a transition zone for nucleation under sonication. For this purpose, a tube with an inner diameter of 1.6 mm and a length of 6 m was winded in a coiled flow inverter (CFI) design and immersed into a cooled ultrasonic bath (37 kHz). The CFI design was chosen for improved radial mixing and narrow residence time distribution, which is also investigated. Amino acid L-alanine dissolved in deionized water is employed in a supersaturation range of 1.10 to 1.46 under quiet and sonicated conditions. Nucleation is non-invasive detected using a flow cell equipped with a microscope and camera. . . . . . Keywords Coiled flow inverter Continuous nucleation Cooling crystallization Minichannel Capillary flow Non-invasive nucleation detection Sonocrystallization Introduction For continuous operation in crystallization, many aspects have to be considered. Clogging should be avoided, while a The continuous operation of small-scale equipment is of ut- narrow crystal size distribution (CSD) is targeted to simplify most importance for accelerated process development [1–4]. It and minimize post-crystallization operations [7, 8]. In addi- is not necessary to switch the operation mode from small scale tion, high yield, high purity, and specific particle morphology batch to the continuous production of bigger quantities, are often required for sufficient product quality. To avoid the smallest quantities can already be continuously produced, preparation of seed crystals some concepts have already been and similar equipment concepts can be transferred more easily developed to integrate the nucleation into the crystallization to larger scales [5]. Furthermore, the continuous mode has process [9–14]. advantages, such as smaller hold-ups, higher product repro- The concept of continuous nucleation by sonication at con- ducibility, and space-time-yield [1, 2, 5, 6]. tinuous flow with good mixing caused by the coiled flow inverter (CFI) design is evaluated in more detail in this con- tribution. The main goal here is to design an apparatus concept Highlights that allows for the non-invasive observation of nucleation due � Sonicated continuous flow for nucleation in coiled flow inverter. to sonication. In the long term, this forms a database which � Nuclei detection non-invasive in continuous flow. will enable a suitable operation for the production of continu- � Determination of two primary nucleation thresholds under sonication, one at the lower limit (transition region starts) and one where nucleation ous seed crystals. occurs in every experiment under given conditions. Mira Schmalenberg and Lena K. Weick contributed equally to this work. Mechanisms of sonocrystallization in solution * Mira Schmalenberg The application of ultrasound for process intensification is a mira.schmalenberg@tu-dortmund.de frequently used and increasingly popular tool [15, 16]. * Norbert Kockmann Ultrasound is defined as sound waves in a frequency range norbert.kockmann@tu-dortmund.de of 16 kHz to 500 MHz [17]. For the transmission of these waves a medium, for example water, is always required BCI Equipment Design, TU Dortmund University, Emil-Figge-Straße 68, 44227 Dortmund, Germany [18]. It is assumed that two phenomena occur simultaneously 304 J Flow Chem (2021) 11:303–319 but can vary strongly depending on the conditions. On the one fragmentations occur for transient and stable cavitation [38]. hand, ultrasound can cause increased molecular movement in With the distinction between transient and stable cavitation, the fluid, which can lead to an increase in the mass transfer as the transient cavitation was observed at 30 kHz and from well as increased mixing [19]. On the other hand, acoustic 41 kHz to 1.14 MHz there is a rather stable cavitation [16]. cavitation bubbles are generated by the impact of ultrasound The collapse intensity of cavitation as well as the active bub- [20]. A detailed description of cavitation bubble formation is ble size and its distribution depend on the ultrasound frequen- given in [21–23]. These cavitation bubbles collapse almost cy, the energy, and the solution itself [16]. Additionally, it has − 1 immediately after they are formed [16, 23], which leads to already been found as rule-of-thumb 35 W∙L (for most sol- local maxima in temperature, while pressure hotspots occur vents operating at room temperature) [15] must be present for during this process [21]. Several theories describe the mecha- cavitation bubbles to form [17, 19]. nism of nucleation by cavitation, but “there is still no consen- sus on the fundamental mechanism of ultrasound action on the Influence parameters of sonocrystallization crystallization process” [16]. The possible processes that are probably influenced by ultrasound during the crystallization Parameters that can be varied and adjusted from the external process are briefly described here and more detail on the pro- environment are the ultrasonic hardware, the frequency, the cesses can be found in [16]. ultrasound exposure, the energy input, and the supersaturation. Concerning the primary homogeneous nucleation, it is dif- Depending on these parameters the induction time, the meta ficult to observe it because there can always be the smallest stable zone width (MZW), the nucleation rate, the polymor- impurities in the system, which would mean heterogonous phism, the yield, the crystal size, and its distribution can be nucleation was detected. A presumed mechanism for homog- influenced. The different ultrasonic system types are ultrasonic enous nucleation is the “segregation theory”, where it is as- baths, which were originally manufactured for equipment sumed that crystal clusters are formed under the effect of an cleaning, probe (horn) systems, and planar transducers [17, acoustic pressure gradient until they result in a larger crystal 21]. The most important advantage of the ultrasonic horn and [16, 24–26]. Other studies show that the acoustic irradiation plate is that the ultrasound can be induced more effectively and leads to overcoming of the critical growth energy [16, 27–29]. with a better focus [12, 39]. The main advantage of the ultra- Most authors postulate that heterogeneous primary nucleation sonic bath is that it is standard laboratory equipment, which is by sonication is present instead of homogeneous primary nu- often already available and has not to be purchased separately. cleation. It is assumed that the reactor surface and smallest The frequency is an important setting for sonocrystallization impurities can reduce the surface energy barrier [16, 30]. because it can affect the bubble dynamics [21, 40], which can Additionally, dissolved gases in the liquid could act as cavi- affect the crystallization, too. The MZW could be reduced in tation nuclei [19]. Therefore, degassing of the solution or fil- comparison to silent conditions for different frequencies (range tering can reduce cavitation bubble formation [16, 31]. of 41–1140 kHz) with constant calorimetric power [41]. The Inversely, investigations on gassing crystallization have higher the used frequency, the less the MZW could be reduced shown that active gassing of a solution can lead to results [41]. If the applied frequency is decreased the formed crystals similar to those of sonocrystallization [32, 33]. become smaller [42]. Nevertheless, Kim and Suslick point out The investigations on cavitation bubble formation using that “the number of cavitating bubbles is not controllable, and sonochemiluminescence by Sarac et al. [34]showed bubble so it is exceedingly difficult to compare different frequencies formation solely using free interfaces generated by slug-flow. due to changes in the number of cavitating bubbles, which are A coiled perfluoroalkoxy alkane (PFA) tube with an inner highly dependent on the specific apparatus used” [21]. Besides diameter of 0.8 mm was placed in an ultrasonic bath the frequency, the intensity is an often-used setting for (27.2 kHz) [34]. No cavitation bubble formation was observed sonocrystallization investigations. In general, it was determined without the introduction of a gaseous phase [34]. that crystals become smaller with increasing US intensity [21]. Regarding secondary nucleation, it is assumed that ultra- Since in this work an ultrasonic bath was used that has sound increases erosion, fracture, and abrasion, resulting in frequency and intensity pre-set, the settings which are more the formation of new nucleation clusters [16, 35]. This in- independently adjustable will be discussed. This includes for crease is due to better mixing by ultrasound, which inter alia example the acoustic irradiation time. It was found that with increases the possibility of crystal collision. However, the increasing sonication time the crystal sizes decrease and be- same mixing effects produced with ultrasound may not pro- come more uniform [21, 43, 44]. duce the same results as mechanical mixing [16, 30, 36, 37]. The energy input is another important setting that should During crystal growth and agglomeration or fracture, the always be taken into account when considering mechanisms “sonofragmentation, sono-deagglomeration and sonocrystallization [17]. In most cases, this parameter results sonoagglomeration” can beobserved, a moredetaileddescrip- from the design of equipment and the hardware used including ultrasound intensity and frequency. However, it can also be set tion can be found in Nalesso et al. [16]. Different types of J Flow Chem (2021) 11:303–319 305 constant to explicitly exclude this influence [41]. Miyasaka was patented, which is equipped with an ultrasonic transducer et al. and Kurotani et al. explicitly investigated the influence and designed with a continuous flow, mainly used for nucle- of the input calorimetric energy in connection with primary ation [54]. nucleation [27, 29]. The calorimetric energy P is determined Narducci et al. were among the first to design a continu- cal − 1 according to Eq. (1) by recording the temperature T change in ously (10–30 mL min ) operated and with 20 kHz ultrasonic the medium as a function of time t and by multiplying the heat processor sonicated stirred vessel (300 mL). In this study, the capacity c of the fluid and the heated mass m. influence of the duration of exposure, the achievement of the steady-state, the crystal habitus, and the particle size distribu- dT tion were already examined. They found out, that the particle P ¼ c m ð1Þ cal p dt size under sonication resulted in smaller crystals [43]. Kudo and Takiyama [55] investigated the generation of For two amino acids, for example, it was found that there is a crystal nuclei for anti-solvent crystallization in segmented local maximum for induction time depending on the ultrasound flow in a metallic and a flexible tube (last one for connection), energy, which in turn is dependent on the degree of supersatura- each with an inner diameter of 2.4 mm. The metallic tube (L = tion and the model system used [27]. The supersaturation itself is 0.5 m) was coiled into a 42 kHz ultrasonic bath [55]. The main also an externally adjustable influencing setting. Jordens et al. results were that monodispersed crystal size distribution showed that the greater the supersaturation, the smaller the crys- (CSD) was created [55]. Further investigations on ultrasonic tals become, whereby an asymptote was detected from supersat- anti-solvent crystallization in milli structured equipment can urationof1.56[42]. Target parameters that are influenced by be found in [56, 57] from the van Gerven group. ultrasound are induction time, the MZW, the nucleation rate, At about the same time the research group of Gruber- the polymorphism, yield, particle size, and its distribution along Woelfler worked on continuous cooling crystallization of with other interactions generated by ultrasonic irradiation [16]. acetylsalicylic acid in ethanol using slug-flow and sonication Ultrasonic irradiation affects the induction time not only − 1 with a flow rate of 6 mL min . The coiled polysiloxane tube for anti-solvent crystallization [44, 45] but also for cooling with an inner diameter of 2 mm and with 1.5 m length was crystallization [27, 46] and usually shortens the induction time inserted into a 35 kHz ultrasonic bath. One point of investiga- probably due to the better micro-scale mixing and the turbu- tion was the nearly constant CSD over the required time lent flow generated by the acoustic cavitation [21]. (25 min). The segmented flow was implemented after the The MZW also decreases under the influence of ultrasound nucleation section and it followed a coiled tubular crystallizer [41, 46, 47]. It was found that higher reduction of MZW was at four different temperature stages. For the whole crystalliza- achieved at low frequencies (41 kHz) [38]. The nucleation rate tion process, different seed loadings were investigated with is presumably increased because the critical excess free energy their influence of a fine dissolution after the ultrasonic bath. is reduced, as the molecule is only surrounded by solution on Furthermore, they analyzed the effect of different ratios of one side and is in contact with the cavitation bubble on the seed-mass and the effect of ultrasound on the product crystals other [21]. Secondary nucleation is also increased because itself. The authors could show that it is possible to design a mixing is improved and the number of collisions is raised [21]. continuously seeded flow-through crystallizer with in-situ So far unknown is the effect on polymorphic forms by seed generation from ultrasound resulting in a constant seed sonication, it is just determined that ultrasound can affect the quantity and quality [11]. polymorphic form [46, 48–50]. A further product-determining A subsequent study, also by Khinast’s research group, factor is the yield, which can be higher with acoustic irradia- investigated the controllability of crystal size in the con- tion because ultrasound reduces induction time so nucleation tinuous tubular crystallizer. They used the same tubes begins earlier [51, 52]. The particle size and its distribution is a (d = 2 mm, polysiloxane) for the coiled crystallizer and very important quantity, also for further processing, so this is a i the same cooling crystallization model system widely studied parameter [27, 40, 42, 43, 51, 53]. Besides, (acetylsalicylic acid, ethanol). However, the seed genera- other interactions can be influenced by ultrasound for example tion via ultrasound was done in a round-bottom flask in interparticle collision, direct interaction between particles, and 500 mL, which was embedded into the ultrasonic bath shockwaves [21, 37]. (35 kHz). Here the supersaturation u was given and the ultrasonic irradiation duration was 60–70 s. The feed so- − 1 Continuous sonocrystallization in capillaries lution flow rate was 22 mL min and the seed suspen- − 1 sion flow rate was 6 mL min , these both flows were Since there are already some studies on nucleation by the mixed before entering the crystallization section (three influence of ultrasound irradiation, the most important ones, thermostat baths each with 5 m coiled tube). At the outlet, considering the setup featured in this work, are presented here. the CSD was evaluated and the crystal size could be con- Already in 2010, an Ultrasonic Oscillatory Baffled Reactor trolled by varying the volume flow rate of the seeds. To 306 J Flow Chem (2021) 11:303–319 avoid clogging, the crystallizer was rinsed with solvent choice and description of the winding type, which will be approximately every 10 minutes [13]. described in more detail below, this and the associated flow Nucleation experiments for different supersaturations in a conditions are simply repeatable. The non-invasive and in-situ range from 1.12 to 1.81 were carried out in another milliscale analysis of the crystals by image analysis ensures that really apparatus, a glass flow cell connected to a PFA tube (d = only crystals are produced in the apparatus and not in the 5 mm), and a stirred vessel, here for the model system para- periphery. The investigations also close the gap that, until cetamol in water. The ultrasonic source was an ultrasound now, crystallization investigations have only been studied transducer (20 kHz), which was mounted on the glass flow for selected supersaturations and not for a merging range. cell or on the stirring tank. The flow rate was selected to be This shows that due to the stochastic process of nucleation, − 1 100 mL min . With ultrasound, nuclei could be generated there is no clearly defined boundary between no crystal for- already from supersaturation of 1.32, and under silent condi- mation and crystal formation. The choice for the continuous tion a supersaturation of 1.87 was necessary [42]. mode of the experiments is not only the basis to construct a Rossi et al. [58] investigated non-invasive with a high- seed crystal unit, but also takes advantage of the fact that many speed microscope cooling sonocrystallization of adipic acid individual batch experiments can be replaced by less contin- in water in diverse supersaturations with a sonoprobe uous runs. (20 kHz). They used a straight process tube made of PFA with an inner diameter of 1 mm and investigated liquid-liquid slug- flow with n-hexane as an inert carrier fluid with a flow rate of Fluid dynamics in a coiled flow inverter − 1 0.3925 mL min [58]. Another research group has recently released a patent, it Many of the discussed research projects used a coiled pipe or comprises the inlet of a substance to be crystallized into a tube for the process flow. It is space-saving and the flow itself tubular crystallizer which is sonicated (20 to 40 kHz) by an exhibits good radial mixing. In capillaries, the velocities are ultrasonic source and includes a temperature control structure usually low with laminar flow, hence, in the macroscopic for cooling the process fluid [59]. Studies on this patent were view, there is low radial mixing in the flow of straight capil- made by Han et al. and Ezeanowi et al. [52, 60]. laries [61]. Han et al. use three ultrasonic baths (35 kHz) connected in To create a macroscopic exchange in the flow, coiled tubes series and equipped with thermostats. In each ultrasonic bath, are used, especially coiled flow inverters (CFI). They are con- a 12 m or 18 m coiled stainless-steel tube with an inner diam- structed of helical segments with a minimum of three turns eter of 4 mm was inserted. Three different substance systems and the segments are linked by 90° bends. The coiled structure namely potassium sulfate, copper sulfate, and phthalic acid itself generates a centrifugal force on the flow, which causes dissolved in water were examined for cooling crystallization. radial mixing, the so-called Dean vortices. Due to the 90° − 1 The mass flow rate was 75 mL min . The residence time bends the direction of the Dean vortices is changing so the dependence, the cooling strategy, the influence of supersatu- radial mixing is improved [62]. ration (1.2, 1.36, 1.54, 1.71, 1.77, and 1.9), and the acoustic Many investigations about coiled tubes and CFIs have been irradiation itself were investigated. No usable crystals could made, especially in characterizing the improved mixing and be produced without sonication. Additionally, no nuclei were the residence time, which could be narrowed by using a CFI formed at a supersaturation of 1.2, and the system was blocked instead of a straight tube [62–67]. Some studies have used the at a supersaturation of 1.9 [52]. CFI design for precipitation [68, 69]orcrystallization Ezanowi et al. investigated a bigger version of the setup in [70–72]. pilot scale with an inner diameter twice as large and approx- The main key numbers used to describe a CFI are briefly imately equal length. They found out that with increasing presented here. The Dean number Dn describes the intensity residence time the CSD decreased and that ultrasound of the Dean vortices formation (Eq. (2)). It is defined as prevented pipe clogging [60]. Reynolds number Re multiplied with the square root of the In most of the studies, few information was provided on ratio of the inner diameter d to the coil diameter d . i c how exactly the tube was coiled. In addition, the analysis was sffiffiffiffiffi usually carried out offline, for example by collecting the prod- d Dn ¼ Re  ð2Þ uct solution in a filter or by using a temperature-controlled collecting-vessel, which both does not display the in-situ mo- The Reynolds number is known as the ratio of inertia force ment. In all the studies described, the analyses have been carried out for selected supersaturations, but it has not been to frictional force (Eq. (3)) and described as mean velocity u multiplied with liquid density ρ and characteristic length, here investigated whether there is a transition or a precise boundary from which nuclei are generated by ultrasound. To fill the the inner diameter, divided by the dynamic viscosity of the liquid η [61]. gaps addressed here, this work was performed. Under exact l J Flow Chem (2021) 11:303–319 307 uρ d l The CFI Design was chosen for its good mixing properties Re ¼ ð3Þ η and the low space consumption caused by the coiled tube. The experimental design is described below. The small inner di- To ensure a narrow residence time distribution in a CFI, ameter (d = 1.6 mm) for the process tube was selected due to − 1 various requirements must be fulfilled. The Dean number has theneed for asufficiently high flow rate (around 16 g min ) to be equal to or higher than three [73], the number of turns to guarantee a homogeneous suspension flow (for the chosen has to be equal to or more than five [73], and the number of model system and 1 w.% solid). Besides, preliminary tests for bends has to be equal to or more than three [62]. Additionally, this combination have already been carried out on a cooling the modified torsion number T , see Eq. 4, has to be equal or crystallizer [72], which in the future will also be connected to higher than 1000, where t is the helix pitch [63, 68]. the plant presented here. After the description of the experimental setup, the analyt- T ¼ Re  ð4Þ ical methods used are presented and the performance regard- ing residence time measurements, nucleation experiments, and calorimetric power is described. In this article, water al- If solids are present in the flow, it is important to avoid ways means deionized water. clogging, a homogeneous suspension flow must be ensured, and no particles should sediment. With an empirical correla- Experimental set‐up tion based on the tube inner diameter, the mean particle size, the particle mass fraction, the suspension flow rate, and the The structure of the plant is divided into three units (Fig. 1). pulsation characteristics (using a peristaltic pump), a homoge- neous suspension flow could be proclaimed. To ensure homo- The first unit consists of the feed and pump unit (FPU), the second is the ultrasonic unit (USU) and the third unit is the geneous suspension flow, the critical Froude number Fr , d,crit product analysis unit (PAU). In keeping with the modular which was developed as an empirical correlation, can be de- concept, this structure would consist of three Process termined to find out, which maximum mean particle size can Equipment Assemblies (PEAs), which could be used as a be transported homogeneously in the tube with a given solid mass fraction. In Eq. (5), the empirical correlation for the single module [75]. The FPU consists of two double-jacketed glass storage critical Froude number is given with model parameters for the switch from moving sediment flow to homogeneous sus- tanks with a filling volume of either 1 or 3.5 L. Here, the feed solution and the solvent are prepared, for which they are pension flow [74]. placed on magnetic stirring plates (MR 3001, Heidolph –4:564 0:717 Instruments GmbH und Co. KG, Germany). The storage tanks Fr ¼ 0:252  Re ðÞ 1–w ð5Þ d;crit s each contain a triangular magnetic stirring rod (PTFE, 50 mm, With the usual definition of the densimetric Froude number Bohlender GmbH, Germany). The tanks are temperature con- trolled with a thermostat (CC304, Huber Kältemaschinenbau (Eq. (6)), the maximum mean particle size x can be deter- mined by converting Eq. (6). The necessary variables are the AG, Germany) and they are connected in series (first the feed storage, second the solution storage). The feed tank contains Reynolds number, the used solid weight fraction w , the mean velocity of the process flow, the liquid density, the density the undersaturated L-alanine water solution, while the solution difference between the solid and the liquid phase Δρ and tank is filled with water for temperature equilibration and sl the acceleration of gravity g [74]. cleaning. The solution is pumped via a peristaltic pump (LabDos Easy-Load, HiTec Zang GmbH, Germany), using a Tygon® tube (inner diameter d 1.6 mm). To ensure that the u ρ i Fr ¼  ð6Þ solution does not contain particles when it enters the process x g Δρ 50;3 sl tube, it is preceded by a filter frit (pore size III, ROBU Experimental section and methods Glasfilter Geräte GmbH, Germany). The process tube is made of fluorinated ethylene propylene (FEP) (Bohlender GmbH, This contribution presents an experimental setup of a coiled Germany) and has an inner diameter d of 1.6 mm. The FPU is tube within an ultrasound bath with an adjacent measuring cell connected with a three-way connector for a pressure sensor below a microscope, which allows in-situ and non-invasive integrated to the third link (A-10, WIKA, Germany) to record observation of the nucleation. To evaluate the flow conditions the pressure loss. The tube from the feed storage to the three- in the coiled tube, the residence time behavior is analyzed. way connector has a length of 1.6 m. Nucleation experiments depending on supersaturation inves- The process tube is coiled in CFI design on a grid basket, tigated whether there is a nucleation threshold or a transition which is available for the used ultrasonic bath. This can be region. Also, particle size distribution, pressure loss, and the seen in Fig. 2 in the upper picture. The total length of the tube time till clogging appears are examined and evaluated. is 6.01 m, including 0.19 m for the inlet and 0.22 m for the 308 J Flow Chem (2021) 11:303–319 Fig. 1 Experimental Set-Up with three units, feed and pump unit (FPU), ultrasonic unit (USU) and product analysis unit (PAU), process flow from left to right outlet. The inlet and outlet of the process tube are insulated to determine the process mass flow rate from the mass with polyethylene foam. The CFI winding on the grid is di- difference with time. vided into six segments with five turns each. The coil diameter Three resistance thermometers (Pt-B-100-2. RÖSSEL- d is 52 mm and the distance between the coils, the pitch, is Messtechnik GmbH, Germany) are used to monitor the temper- about 3.6 mm. The coiled tube on the grid basket is embedded ature with one in the feed or solution storage, one in the ultra- in the ultrasonic bath (Elmasonic S30H, Elma Schmidbauer sonic bath, and one in the direct environment. Before the ex- GmbH, Germany) filled with deionized water. The bath has a perimental runs, all temperature sensors were calibrated by a filling volume of 2.75 L [76], whereas it is only filled with two-point calibration procedure with two thermostats. The inlet 2.19 L so that the coiled tube is always fully covered with and outlet temperature of the process medium are recorded via water. Moreover, the tube does not touch the ground or walls sealed thermocouples (Typ K, 1 mm diameter, OMEGA of the ultrasonic bath. According to the manufacturer, the Engineering, United States). From preliminary tests, the follow- ultrasonic bath has a frequency of 37 kHz, the total power ing correction Eq. 7 of the measured temperatures on the tube, consumption is 280 W, while 80 W being the effective ultra- and the actual temperature could be determined. This correction sonic power [76]. To ensure a constant temperature in the is valid in the temperature range 10 °C to 60 °C. ultrasonic bath for the nucleation experiments, it is connected to a cryostat (Pilot ONE ministat 125, Huber Kältemaschinenbau T ½ °C ¼ 1:1774  T ½ °C −3:9049 ð7Þ fluid surface AG, Germany). The internal pump rate is set to 3000 rpm. For the process control system, a LabManager® (HiTec The process tube, which comes out of the USU, is con- Zang GmbH, Germany) is used, data evaluation is enabled nected to a FEP tube (same inner diameter) of 0.76 m in LabVision® (HiTec Zang GmbH, Germany). The data is length and leads to the PAU. The next part of the tube recorded every second. (1.2 m) leads from the measuring cell (Fig. 2 bottom) into the sampling tank. The section of the tubing located inside the measuring flow cell has a length of 325 mm. The mea- Measurements suring flow cell in combination with a microscope (Bresser Science ADL 601P, Bresser GmbH, Germany), a photo Pressure loss camera (Nikon Z6, Nikon GmbH, Japan) and an image evaluation tool [77] based on the algorithms of Borchert In addition to temperature monitoring, the pressure loss Δp and Sundmacher[78]and Huoetal. [79] enables a non- for the experiments is measured by the pressure sensor (A-10, invasive in-line measurement of the crystals produced. WIKA, Germany). The pressure loss can also be correlated to For further details refer to [77]. The processed fluid is the mean flow velocity u according to Eq. (8), where f is the finally collected in a collecting container. This is located loss coefficient, ρ the fluid density, and L the length of the on a scale (Kern 572, KERN & SOHN GmbH, Germany) tube[61]. J Flow Chem (2021) 11:303–319 309 Fig. 2 Process tube coiled on a grid basket in CFI design (upper), measurement cell with process tube under the microscope (bottom) special temperature control of the measurement cell was nec- ρL essary although this is possible [77]. Δp ¼ f u ð8Þ 2d The photos taken are evaluated using a routine, which is based on the algorithms by [78]and [79]. This semi-automatic A suitable correlation for the friction factor f for coiled image processing routine determines the circular area equiva- tubes is Eq. (9)[80] and has already been used successfully lent diameter, using the projection surface of the crystals [77]. in other studies with tubes coiled in CFI design [63, 72]. It is composed of the pipe friction coefficient for straight pipes f , with 64/Re for laminar flow, and the Dn number. Concentration determination via gravimetric method 0:603 f ¼ f 1 þ 0:0456  Dn ð9Þ a 0 A gravimetric method is used to determine and check the Another correlation for the friction factor f for coiled tubes b dissolved amount of the model substance (L-alanine). and laminar flow is Eq. (10)[81]. Therefore, the quotient of the mass of the target component i and the mass of the solution sol is determined by Eq. (11). f ¼ f 1 þ 0:033 ðlog ðDnÞÞ ð10Þ b 0 10 m m  m i dried empty w ¼ ¼ ð11Þ Both correlations are used and compared with each other in m m  m sol sample empty order to enable better predictions to be made for future operations. An empty (and labeled) 1.5 mL reaction vessel is weighed with a scale (XA 205 Dual Range, Mettler Toldeo, US) to determine m .To get m a syringe with pre-filter (cel- empty sample Particle measurement via image analysis lulose acetate membrane, pore size 0.2 µm, VWR International, US) is filled with the solution sample (about To investigate the formed crystals, an image analysis protocol, 1 mL) and injected into a reaction vessel, and weighed again. which was already mentioned in the section above, was used. After drying the sample in a vacuum drying oven (Memmert The special property here is that the crystals can be detected VO400, Memmert GmbH + Co. KG, Germany) at 60 °C and non-invasive without large equipment effort. As a major com- with stepwise reduction of the pressure (1 bar, 600 mbar, ponent, a microscope is equipped with a camera, and a mea- 300 mbar), it will be weighed again to get m .Each sample dried surement cell is placed below. The process tube is clamped is dried for a minimum of 14 days and is considered un- through the measuring cell, which is out of glass and filled changed if the sample does not change for more than with water to observe the crystals distortion-free. Because the 0.001 g, almost the same procedure has already been success- system was mostly cooled down to room temperature, no fully carried out in [74]. 310 J Flow Chem (2021) 11:303–319 Experimental procedures method. This number (Eq. 17) describes the relationship between convective mass transport and axial mixing, First of all, the experimental procedure to determine the which includes the axial dispersion coefficient. It is also residence time of the liquid phase in the USU is described, used to indicate if axial mixing is negligible compared to followed by explaining the performance of nucleation exper- convection, i.e. the residence time behavior is comparable iments with and without ultrasound. The most important with an ideal plug flow (Bo > 100) [74, 82]. parameters are also defined here. Finally, the procedure is presented for determining the dissipated energy by the ultra- t ¼ ∫ ðÞ 1−FtðÞ dt ð14Þ t¼0 sonic bath. θ ¼ ð15Þ Residence time behavior pffiffiffiffiffiffi 1 Bo 1−θ pffiffiffiffi FðÞ θ ¼ 1–erf  ð16Þ 2 2 θ To determine the actual liquid residence time t and the distri- bution sum function F(t), the step-response method was used. uL For this purpose, deionized water (from solution storage) is Bo ¼ ð17Þ ax first introduced into the entire apparatus (USU). The inlet in the FPU was already filled with tracer solution (from feed For the description of the residence time distribution in a storage) up to the second three-way valve (see Fig. 1)in order CFI, it could already be shown that the dispersion model is to measure only the residence time distribution in the USU well suited [72, 74, 83]. (pressure sensor was not used here). As a tracer, a thiosulfate − 1 solution with a concentration of 0.2–0.4 mg·g water (not changing the parameters of water significantly) [73]is used Nucleation experiments and detected by a UV-Vis detector (Evolution 201, Thermo Fisher Scientific Inc., US). Instead of the microscope in the An amino acid (L-alanine, purity ≥ 99.6%, Evonik Rexim PAU, the UV-Vis detector with a measuring flow-through (Nanning) Pharmaceutical Co., Ltd., China) and water were cuvette (170700-0.2-40, Hellma Analytics, Germany) is used. used for the nucleation experiments with and without ultra- − 1 The mass flow rate is set to 16.5 ± 0.8 g min .To inves- sound. Based on the empirically determined solubility curve tigate the influence of the US on the residence time the exper- (Eq. (18)) according to Wohlgemuth et al. [84], the L-alanine/ iments were done three times without ultrasound and four water solution was prepared for the desired saturation temper- times with ultrasound. The ultrasonic bath was cooled to room ature T . temperature with the connected cryostat. −3 * For calculating the hydrodynamic residence time τ * ðÞ 9:0849210 T ½ °C w½ g  g ¼ 0:112381  e ð18Þ sol (Eq. (12)), the known geometry of the tube (d = 1.6 mm, L = 6 m) for the internal volume V was considered as well as the The preparation of the solution in the FPU was done one liquid density ρ , which is almost identical to that of water. day before the experiments started. Therefore, the feed tank The values for water are taken from [81]. temperature (TI_00) was set to 5 K higher than the chosen Vρ l saturation temperature (T = 50°C or60°C).Thisensures τ ¼ ð12Þ m that the crystals are all dissolved and that no nuclei were al- ready present in the feed tank. Hence, the solution is not sig- The distribution sum curve is defined as Eq. 13, with con- nificantly oversaturated before entering the ultrasonic bath. centration w at time t in dependence of the tracer concentration The temperature of the ultrasonic bath was set for each exper- w . iment, depending on the target supersaturation, for example to 20 °C (ΔT = 30 K) for a supersaturation of 1.3. To determine wtðÞ FtðÞ ¼ ð13Þ the supersaturation from temperature S , Eq. (19) was chosen, using the assumption that the activity of the dissolved compo- The actual residence time t (Eq. 14) was determined by nent behaves ideally [84]. Since this is only a value calculated integration of the step-response, using the trapezoidal from the measured temperatures, the gravimetric supersatura- rule. The use of the dimensionless time θ (Eq. 15)isused tion was determined additionally. Therefore, instead of the to compare each experiment. With the use of the disper- calculated solved fraction of L-alanine (using Eq. 18), the sion model Eq. 16, which allows to consider the axial gravimetrical measured fraction of the feed tank w was used. back mixing, the Bodenstein number Bo was determined The determined supersaturation is labeled as S in the follow- by fitting the experimental data using the least square ing (Eq. (20)). J Flow Chem (2021) 11:303–319 311 * * w T could be easily detected whether nucleation has occurred in S ¼ ð19Þ wTðÞ the case of different supersaturations. out S ¼ ð20Þ G Residence time behavior wðÞ T out The residence time distribution of the CFI in the USU is in- The sonicated and silent nucleation experiments are done vestigated for two reasons. On the one hand, the apparatus is in a supersaturation range of 1.10 to 1.36 for a saturation to be characterized entirely and the hypothesis of a narrow temperature of 50 °C and some extra investigations are made residence time will be confirmed by the CFI design. On the with an inlet saturation temperature of 60 °C for the supersat- other hand, the observations regarding acoustic streaming, uration 1.38, 1.42, and 1.46. The supersaturation was adjusted mentioned below, made by Valitov et al. [86] will be verified. via the end temperature. Before each experiment, the temper- The residence time behavior of the tube in CFI design is ature profile runs in with water until it was stationary (for shown in Fig. 3. Here the residence time sum curve under minimum of 20 min) before switching to the undersaturated quiet and sonicated conditions is plotted as a function of di- solution and starting the nucleation experiments, where each mensionless time, each with the experimental data (symbols) experiment is either run till clogging or 20 min if no crystals and fitted to the dispersion model (green line). It can be seen were visible. that the residence time distribution curve is very narrow and similar to an ideal tube. No optical differences between the Calorimetric energy input silent and sonicated curve can be seen. The most important parameters are listed in Table 1. As already mentioned, Eq. (1) is used to determine the energy − 1 The mass flow rate of 16.5 ± 0.8 g min could be kept input by the ultrasonic bath. This method assumes that the almost constant during the different tests. This allows a good whole power entering the solution is dissipated as heat [42, comparison between the silent and sonicated tests. Using the 85]. Therefore, two setups were used. Case 1 only includes the mass flow rate, the Reynolds number, Dean number, and USU and the 2nd case is the same setup as described in the modified torsion number can be determined to 238, 42, and experimental setup section. In both cases, the ultrasonic bath 10,785 respectively. These values show that the flow is in the itself was not temperature controlled via the cryostat. laminar range and the CFI design fulfills the requirements for The environmental temperature, the ultrasonic bath temper- a narrow residence time distribution. ature, and in the second case also the in- and outlet tempera- The hydrodynamic residence time is slightly shorter than ture of the process flow were measured. In every experiment, the real evaluated residence times. This could indicate that the temperature change was recorded in dependency of the there might be a small death zone; probably it is the flow- time. For every experiment, first, the ultrasonic bath tempera- through cuvette in the UV-Vis detector itself. The actual res- ture is prepared beneath or at room temperature. Then, the idence times for silent and sonicated conditions are also very ultrasonic bath was turned on. For the second case, a process similar, both within the error tolerance, but the sonicated one flow (water) with room temperature through the CFI tube is slightly smaller. The calculated Bodenstein numbers are − 1 (16.7 ± 0.1 g min ) is turned on. Each of these cases were higher than 100 for both cases, hence, the residence time be- measured four times and for a minimum time of 20 minutes. havior is similar to an ideal plug flow. Additionally, the one The performance of the ultrasonic bath was recorded for for the sonicated conditions is a bit higher than for the silent, comparability with other studies. The measurements showed a but in the range of the standard deviation, too. Therefore, it − 1 temperature change of 0.007 K·s . According to Eq. 1,this can be concluded, that the sonication has no significant effect results in an energy of 0.35 W applied to the volume of the on the residence time distribution if the CFI design is chosen. tube in the ultrasonic bath using the heat capacity of water Valitov et al. [86] investigated and evaluated the influence [81]. of the ultrasound (horn, 20 kHz) on the flow of a straight FEP tubing with inner diameters of 1.55 mm and 3.2 mm for a sonication time of 0.3, 1.0, and 3.0 s. They assumed that Results and discussion boundary layer streaming could be neglected because FEP as a material is acoustically transparent and boundary layer To verify a narrow residence time distribution in the CFI the streaming in millifluidic devices tends to be visible only at residence time behavior was investigated. Furthermore, it was higher frequencies (> 1 MHz). Furthermore, they considered evaluated whether it is possible to investigate nucleation in the and evaluated the Eckart streaming and microjetting, but for described setup with and without sonication so that the MZW the small capillary (1.55 mm), which was also used in this of primary nucleation can also be determined for continuously research, the influence of these two flow phenomena, conse- operated crystallizers. By non-invasive particle observation, it quently acoustic flow, on the residence time behavior could be 312 J Flow Chem (2021) 11:303–319 Fig. 3 Residence time sum curve, in dependency of the dimensionless residence time, without periphery for the CFI in the USU for silent (left) and son- icated (right) conditions neglected. This means that the pure convection model can be determined from the difference of the saturation and outlet applied in the work of Valitov et al. to describe the residence temperature (ΔT) or from the inlet concentration and the out- − 1 time behavior of the sonicated straight tube of 1.55 mm inner let temperature. The mass flow rate was set to 17 g min . diameter. For the bigger capillary and longer sonication time Figure 4 shows the individually obtained supersaturations. (3.0 s) the influence of the acoustic streaming could be shown, A distinction is made between the sonicated and silent condi- so the dispersion flow model fitted better than the pure con- tions, and some of the individual experiments could be aver- vection model and indicated the presence of axial back mixing aged into one sample because of their similarity. The results due to the sonication, so the residence time behavior was are divided into three categories. Either no crystals were vis- influenced in this case [86]. ible the whole time, crystals were occasionally detected in the The results of residence time behavior in the CFI confirms transition range, or nucleation crystals were continuously vis- the investigations of Valito et al. [86] and would suggest that ible. Accordingly, crystal nuclei were repeatedly seen in the the acoustic flow in these capillaries (d 1.6 mm) can be transition area during the test, but there were also longer pe- neglected. Nevertheless, the small deviations of the actual riods (minutes) in which not a single crystal was seen under residence time and the Bodenstein number between the silent the microscope. If none of the trials would be assigned to the and sonicated conditions could be a hint that for small capil- transition range, there would be a supersaturation threshold laries and longer sonication times a small effect of the acoustic above which crystals could always be generated. If there is a streaming is present. transition range, then there are two boundaries: the first be- tween where no crystals are formed and stochastically crystals are formed and the second between stochastically crystals are Nucleation experiments formed and crystals are visible the whole time. Assuming that the nucleation behavior of a substance sys- The nucleation experiments were done under silent and soni- tem would be the same in a batch apparatus as in a cated conditions. To examine whether the nucleation in CFI under ultrasound exposure has a sharp limit or rather a transi- tion area, most of the supersaturations were in the area of 1.10 to 1.46. The supersaturation could not be completely reduced in any of the tests. Accordingly, the indicated supersaturation here always indicated the calculated supersaturation, which is Table 1 Key figures for the residence time of the CFI silent −1 m gmin T [°C] ts½ [s] Bo [-] 16.5 ± 0.8 23.3 ± 0.5 43.8 45.3 ± 0.7 161.5 ± 4.8 sonicated −1 m gmin T [°C] ts½ [s] Bo [-] Fig. 4 Nucleation in the USU under sonicated and silent conditions for different supersaturations; samples A – D are averaged values from sev- 16.5 ± 0.8 23.3 ± 0.5 43.8 44.8 ± 0.1 167.1 ± 6.9 eral experiments at almost the same supersaturation J Flow Chem (2021) 11:303–319 313 continuously flowed crystallizer, the following experimental cavitation bubbles are size distributed, which in turn leads effort between the different modes would result. Each exper- to size distribution of crystal nuclei. imental point is run over at least 14 to 20 minutes, with a When interpreting the results, it should be considered that residence time of 45 s. This results in about 18 to 26 experi- the influence of a different residence time and, consequently, a mental runs. If this were done within a batch experiment, the longer possible induction time has not been investigated, be- cooling rates would have to be run in again 18 to 26 times, and cause of the given boundary conditions (mass flow rate and after 45 s each time it would have to be determined whether manageable coiling length of the tube). If the residence time in nucleation occurred or not. This shows how much experimen- the ultrasonic bath could have been extended by a longer tube, tal effort can be saved by this newly introduced methodology. Fig. 4 would probably have changed that crystals also have First, for the nucleation experiments without ultrasound, been seen at lower supersaturations or the transition range i.e. under quiet conditions, it can be seen that no crystals were would have been smaller, since the possible induction time visible, neither for the smaller supersaturations in the range of could be longer. 1.10 to 1.33 (ΔT = 10 K to 34.1 K) and a saturated starting Since eight experiments were performed with approximate- solution at T = 50 °C (filled grey circles) nor for the higher ly the same supersaturation of 1.25, these experiments were supersaturations 1.42 and 1.46 (ΔT=34 K to41.3K)witha combined as sample A, and shown as a cross with a green saturated starting solution at T = 60 °C (empty circles). background (Fig. 4). This means that the S and S of these T G For a continuous stirred tank, which is especially construct- tests are averaged as well as their mass flow rate, ambient ed for nucleation of the system used, it is known from the temperature T and the determined and calculated pres- ambient, literature [10] that nucleation takes place from subcooling of sure loss. These values are listed in Table 2. This averaged about 2 K (46 °C to 44 °C, or supersaturation of 1.02) with an sample A consists of one experiment, in which no crystals induction time of 1.68 min. It should also be considered that were visible, three experiments, in which crystals were par- for continuously operated crystallizers, not only the MZW can tially visible, and four experiments with sustained nucleation. be considered, but also the residence time. Thus, the induction This leads to the conclusion that for the selected material sys- time must be considered and the MZW should therefore be tem and the apparatus arrangement the transition to a most called a primary nucleation threshold (PNT) [10]. Another probable formation of crystal nuclei is present from supersat- study investigated the MZW for different saturation and nu- uration of 1.249 ± 0.005. cleation temperatures for L-alanine/water in a batch stirred For other supersaturations in the nucleation range, experi- vessel [87]. Extrapolation of these data results in a required ments could be combined into one to compare these points supersaturation of about 1.4 for an induction time of around with each other in a more targeted way. For this purpose, at 45 s. least three experiments were averaged, which are also listed in The results of the silent nucleation tests suggest that good Table 2 with their associated process variables. In the diagram, mixing, produced by the flow in the CFI design, is not suffi- these samples B – D are also shown as blue symbols. In cient to produce nucleation at lower supersaturations because summary, it is shown as described before that lower supersat- even at a supersaturation of 1.46 there were no nuclei visible. urations are required for nucleation by sonication than with- Presumably, the PNT for silent conditions is even wider. out. Further, it was found that there is a transition area, in However, this was not considered further, since Eder et al. which nucleation takes place in the selected apparatus and [11] already found that for the generation of seed crystals in chosen model system. For the selected conditions it is found a tube without ultrasound, the system usually clogs. that the PNT starts at a supersaturation of 1.2 and above a Every single experiment under sonication is marked as x supersaturation of 1.26 crystallization always occurs. and sorted into the categories for the different In comparison to the data from the literature [42, 52, 58], (temperature) supersaturations in Fig. 4. Considering this these results could indicate that using lower frequencies in the arrangement, it was found that in a supersaturation range of rage of 20–40 kHz, a critical nucleation threshold in a super- 1.20–1.25 individual experiments are sorted into all three saturation range of 1.25 to 1.3 is present. To verify this state- categories. Accordingly, PNT was significantly reduced by ment, more investigations should be carried out with the same sonication of the tube. No special effect could be deter- equipment and the same energy input for many different sub- mined regarding the experimental days or the ambient tem- stance systems. perature. Therefore, the results show that under the select- The results also confirm the hypothesis that nucleation is − 1 ed process conditions of flow rate (17 g min )and used caused by the cavitation bubble rather than by the acoustic experimental setup there is no precisely defined supersat- flow, because the frequency range used is known for pro- uration threshold above which nucleation is generated by nounced cavitation bubble formation [16]. Additionally, ultrasound, but rather a transition region. This could be Valitov et al. [86] showed that acoustic flow is rather negligi- because both nucleation and the formation of cavitation ble for this capillary size, and even with an already very well- mixed system, generated by the tube design, the PNT could bubbles are stochastic processes. It is also known that 314 J Flow Chem (2021) 11:303–319 Table 2 Averaged nucleation experiments in the USU under sonicated conditions for different supersaturations −1 number of experimental S [-] m gmin T status Δp Δp [mbar] Δp [mbar] Δp ambient exp calc calc ̅calc runs [°C] [mbar] with f with f [mbar] a b A 8 1.249±0.005 17.2±0.3 23.5±1.0 transition 195.5±6.8 181.1 214.7 197.9 range B 3 1.270±0.005 17.3±0.4 23.5±0.3 nucleation 199.5±6.9 184.9 218.8 201.9 C 4 1.284±0.005 17.3±0.3 22.9±0.9 nucleation 206.5±6.6 188.9 222.5 205.7 D 4 1.298±0.003 17.3±0.3 23.2±0.9 nucleation 213.1±8.7 190.6 224.5 207.6 not be reduced. However, no cavitation bubble formation Pressure drop and clogging could be observed in a smaller capillary (d = 0.8 mm) without air slugs in the investigations of Sarac et al. [34]. In order to The experimentally determined pressure loss Δp is com- exp ensure that cavitation bubble formation is responsible for nu- pared to the calculated one. Figure 6 shows the measured cleation, it should be investigated if cavitation bubbles are pressure loss (filled rhombus), the calculated pressure loss formed in the capillary size of 1.6 mm. coefficient f (squares) and f (triangles) and the averaged a b pressure loss from both (plus sign) over the Reynolds number. It can be seen that the actual pressure loss with loss coefficient Crystal size distribution f calculated is too low and f is overestimated. Averaging the a b two calculated pressure losses gives a better prediction of the To compare the samples A – D, the relating CSD was also pressure loss for the range investigated here. This calculation investigated. In addition, another single experiment with su- can help to follow an automated operation by starting a flush- persaturation of 1.38 was added for evaluation. Photos were ing cycle in case of deviating pressure loss. taken at random times during the experiments, at least with a For most of the nucleation experiments, the time of clog- time interval of three seconds to ensure that no crystals are ging was also recorded, shown in Fig. 7. The shortest test photographed twice. Crystals with a diameter of less than duration was thus about 14 min and a maximum of 35 min. 10 µm were not included in the evaluation, as the image anal- Dependence of the supersaturation on the clogging time t clog ysis could not determine whether these were actually formed could not be detected, because both extrema could be found at crystal nuclei or only image artifacts. These random photos the same supersaturation. The mean time until clogging was were evaluated using the image evaluation tool described 22 min. Although a clogging is usually visible due to the above [77]. The results are shown in Fig. 5. The determined transparent tube, no location could be identified where clog- supersaturation is plotted as dependence of the diameter. This ging has repeatedly occurred. Additionally, it was not possible is the circular equivalent diameter. The number of optically to identify excessive solid loads, which could be the cause of evaluated particles is listed above each CSD. The crystal size clogging. It is assumed that the small particles (< 90 µm), distributions are drawn in classes (black points, with a class which tend to agglomerate more frequently [88], have caused width of 10 µm) and as boxplots. the clogging, although the homogeneous flow conditions are Based on the CSD results, no clear influence between su- fulfilled. Besides, the already formed crystal nuclei can lead to persaturation and the influence on crystal size can be identi- a strengthening of secondary nucleation, because not enough fied. For the lower supersaturations (< 1.3) a particularly larg- surface for the crystals to grow is available and, instead, sec- er amount of fine grain (< 25 µm) was detected. This can also ondary nucleation occurs, which leads to clogging due to the be seen in the small values for x . The mean crystal size is relatively high supersaturation. 10,0 between 24 and 49 µm. At the higher supersaturation single A similar experimental setup to ours can be found in the experiment, the particles are bigger and the x is 62 µm. work of Furuta et al. [89] with a pH swing crystallization. 50,0 However, since a very different number of crystals were eval- Here, a coiled tube made of PFA (d =2 mm)and alength uated in each experiment, an interpretation of the dependence of 20–40 m was inserted into a temperature-controlled ultra- − 1 between supersaturation and crystal size is not statistically sonic bath (40 kHz). The flow rate was set to 50 mL min . validated. For this reason, the only conclusion to be drawn However, the influence of ultrasound on the formation of nu- from these results is that crystals with a size of 10–150 µm clei was not investigated by the authors. The ultrasonic irradi- were produced, but the majority of them are below 100 µm. ation was only used to avoid clogging [89]. Therefore, the created crystals are in the same range of crystal In other research groups, too, observations have been made size as they were created by Eder et al. [11] with a different on clogging. For example, Eder et al. [11] could not achieve model system, but in most similar conditions. stable seed crystal formation without ultrasound. Likely, J Flow Chem (2021) 11:303–319 315 Fig. 5 Crystal size distribution, evaluated with image analysis, in dependence of the supersaturation (for averaged nucleation experiments (A-D) and one addi- tional experiment with higher su- persaturation (1.38)) direct clogging would also have occurred in our case if even here that the rinse cycles were scheduled after 10 min [13], i.e. higher supersaturations had been applied to produce crystals in a similar time range as our shortest run till clogging was without ultrasound. In the investigations of Han et al. [52]a observed. If stable nucleation operation without clogging is feasible, stable operation was defined, if 3 to 4 residence times guaranteed, the USU presented here can be used as a seed no clogging occurred. According to this definition, all exper- crystal generation unit for continuous crystallization equip- iments presented here would have run stably. However, this ment as known from [71, 72, 90–93]. cannot be compared directly because the diameter chosen by Han et al. is 4 mm [52]. But, their average velocity is in a − 1 comparable range from 8 to 16 cm s [52]toours ofaround − 1 13 cm s . They also found that supersaturations that are Conclusions punctual high and not uniformly distributed lead to clogging. In the case of Ezeanowi et al. [60], no clogging was observed Cooling crystallization was investigated within a capillary − 1 at similar average velocities of 5–9cm s and an internal coiled flow inverter (CFI) with a tube diameter of 1.6 mm diameter of 250 mm. The observed duration per experiment embedded in an ultrasonic bath for nucleation in a continuous was also 3–4 residence times [60]. flow with different supersaturations. For this purpose, the out- In combination with the results determined here, it can be let of the test facility has been equipped with a non-invasive assumed that clogging can be prevented with even higher flow cell under a microscope with a conventional photo cam- suspension flow, despite the previously mentioned process era. The CFI design ensures a very narrow residence time condition of homogeneous suspension. As an alternative to behavior of the liquid phase. At a mass flow rate of 16.5 g − 1 avoid clogging problems, a feed mixture with a solution and min , a Bodenstein number of approx. 161–167 was obtain- seed suspension can be used, or rinse cycles can be imple- ed with a residence time in the ultrasonic bath of 45 s. Only mented as Besenhard et al. [13] did. It is interesting to note very slight differences between the residence time under Fig. 7 Time of clogging after starting the process solution for the Fig. 6 Measured and calculated pressure loss in dependence of the respective supersaturation Reynolds number 316 J Flow Chem (2021) 11:303–319 Competing financial interest The authors declare no competing finan- sonication and quiet conditions could be detected which could cial interest. indicate an influence of the acoustic streaming. Under silent conditions, no nucleation was observed in the Abbreviations CFI, coiled flow inverter; CSD, crystal size distribution; investigated range of supersaturation from 1.10 to 1.46 for the FPU, feed andpumpunit; FEP, fluorinated ethylene propylene; MZW, model system L-alanine/water. The first nuclei could be de- metastable zone width; PFA, perfluoroalkoxy alkane; PNT, primary nu- cleation threshold; PEAs, Process Equipment Assemblies; PAU, product tected under sonication and supersaturation of 1.2. The results analysis unit; US, ultrasound; USU, ultrasonic unit of the nucleation experiments show that under the selected conditions (ultrasonic bath, CFI designed tube, L-alanine/wa- Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adap- ter, and residence time 45 s) the primary nucleation threshold tation, distribution and reproduction in any medium or format, as long as (PNT) could be reduced by ultrasound as expected. you give appropriate credit to the original author(s) and the source, pro- Additionally, no rigid threshold for nucleation could be found, vide a link to the Creative Commons licence, and indicate if changes were but a nucleation range was identified in the investigated setup. made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a Nucleation was detected for supersaturations larger than 1.25. credit line to the material. If material is not included in the article's The transition range indicates that the two stochastic processes Creative Commons licence and your intended use is not permitted by of cavitation bubble formation and nucleation must be suffi- statutory regulation or exceeds the permitted use, you will need to obtain ciently pronounced to ensure nucleation. No influence of the permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. set supersaturation could be determined on the crystal size distribution or the time until clogging. The minimum time until clogging occurred was 14 minutes. The crystal size is in a range from 10 to 150 µm, where the most are from the size References in the range of 24–49 µm after a residence time of 45 s. Overall, the investigated setup is suitable for the rapid 1 Ma Y, Wu S, Macaringue EGJ et al (2020) Recent progress in detection of the influence of supersaturation on material continuous crystallization of pharmaceutical products: precise prep- systems by sonication. This setup consisting of coiled aration and control. Org Process Res Dev. https://doi.org/10.1021/ acs.oprd.9b00362 polymer capillary within an ultrasound bath can easily be 2 Yazdanpanah N, Nagy ZK (2020) The Handbook of Continuous set up with existing laboratory equipment. The developed Crystallization. Royal Society of Chemistry, Cambridge measuring cell from a glass jacket, optical microscope, and 3 Melches C, Plate H, Schürhoff J et al (2020) The steps from digital camera could even enable to identify influences on batchwise to continuous crystallization for a fine chemical: a case crystal form quickly and easily by ultrasound. In addition, study. Crystals 10:542. https://doi.org/10.3390/cryst10060542 the ultrasonic unit (USU) can be used as a seed crystal 4 Zhang D, Xu S, Du S et al (2017) Progress of pharmaceutical con- tinuous crystallization. Engineering 3:354–364. https://doi.org/10. production unit and in combination with rinse protocols 1016/J.ENG.2017.03.023 for a combination of continuous crystallization for the 5 Bieringer T, Buchholz S, Kockmann N (2013) Future production small-scale with other crystal growth units. concepts in the chemical industry: modular - small-scale - continu- ous. Chem Eng Technol 36:900–910. https://doi.org/10.1002/ceat. Acknowledgements The authors would like to thank Prof. Dr. Gerhard Schembecker and Dr. Kerstin Wohlgemuth (TU Dortmund University, 6 Schaber SD, Gerogiorgis DI, Ramachandran R et al (2011) BCI Laboratory of Plant and Process Design) for their inspiring discus- Economic analysis of integrated continuous and batch pharmaceu- sions on the investigations. We would also like to thank the reviewer for tical manufacturing: a case study. Ind Eng Chem Res 50:10083– the diligent review and helpful comments. Furthermore, we thank our 10092. https://doi.org/10.1021/ie2006752 technician Carsten Schrömges for the technical support. 7 Sun (2010) Particle size specifications for solid oral dosage forms: a regulatory perspective. https://www.americanpharma Author contributions The manuscript was written through contributions ceuticalreview.com/Featured-Articles/36779-Particle-Size- of all authors. All authors have given approval to the final version of the Specifications-for-Solid-Oral-Dosage-Forms-A-Regulatory- manuscript. Perspective/. Accessed 7 Oct 2020 8 Jong EJ de (1982) Entwicklung von Kristallisatoren. Chem Ing Tech 54:193–202. https://doi.org/10.1002/cite.330540303 Funding Open Access funding enabled and organized by Projekt DEAL. This research was funded by the German Federal Ministry of Economic 9 Hohmann L (2019) Process and equipment design for small-scale Affairs and Energy (BMWi) and the Project Management Jülich (PtJ) as continuous crystallization, 1. Auflage. Schriftenreihe part of the ENPRO2.0 initiative (Ref. no. 03ET1528A). Additionally, we Apparatedesign, Band 11. Dr. Hut, München acknowledge financial support by Deutsche Forschungsgemeinschaft and 10 Lührmann M-C, Timmermann J, Schembecker G et al (2018) Technische Universität Dortmund/TU Dortmund University within the Enhanced product quality control through separation of crystalliza- funding programme Open Access Publishing. tion phenomena in a four-stage MSMPR cascade. Cryst Growth Des 18:7323–7334. https://doi.org/10.1021/acs.cgd.8b00941 11 Eder RJP, Schrank S, Besenhard MO et al (2012) Continuous Compliance with ethical standards Sonocrystallization of Acetylsalicylic Acid (ASA): control of crystal size. Cryst Growth Des 12:4733–4738. https://doi.org/10.1021/ Conflicts of interest/Competing interests No conflict of interest cg201567y J Flow Chem (2021) 11:303–319 317 12 Jiang M, Papageorgiou CD, Waetzig J et al (2015) Indirect 31 Lee J (2016) Handbook of ultrasonics and sonochemistry: ultrasonication in continuous slug-flow crystallization. Cryst Importance of Sonication and Solution Conditions on the Acoustic Growth Des 15:2486–2492. https://doi.org/10.1021/acs.cgd. Cavitation Activity. Springer, Singapore 5b00263 32 Wohlgemuth K, Ruether F, Schembecker G (2010) 13 Besenhard MO, Neugebauer P, Ho C-D et al (2015) Crystal size Sonocrystallization and crystallization with gassing of adipic acid. control in a continuous tubular crystallizer. Cryst Growth Des 15: Chem Eng Sci 65:1016–1027. https://doi.org/10.1016/j.ces.2009. 1683–1691. https://doi.org/10.1021/cg501637m 09.055 14 Wong SY, Cui Y, Myerson AS (2013) Contact secondary nucle- 33 Wohlgemuth K, Kordylla A, Ruether F et al (2009) Experimental ation as a means of creating seeds for continuous tubular crystal- study of the effect of bubbles on nucleation during batch cooling lizers. Cryst Growth Des 13:2514–2521. https://doi.org/10.1021/ crystallization. Chem Eng Sci 64:4155–4163. https://doi.org/10. cg4002303 1016/j.ces.2009.06.041 15 Ruecroft G, Hipkiss D, Ly T et al (2005) Sonocrystallization: the use 34 Sarac BE, Stephens DS, Eisener J et al (2020) Cavitation bubble of ultrasound for improved industrial crystallization. Org Process dynamics and sonochemiluminescence activity inside sonicated Res Dev 9:923–932. https://doi.org/10.1021/op050109x submerged flow tubes. Chem Eng Process Process Intensif 150: 16 Nalesso S, Bussemaker MJ, Sear RP et al (2019) A review on pos- 107872. https://doi.org/10.1016/j.cep.2020.107872 sible mechanisms of sonocrystallisation in solution. Ultrason 35 Hem SL, Skauen DM, Beal HM (1967) Mechanism of crystalliza- Sonochem 57:125–138. https://doi.org/10.1016/j.ultsonch.2019.04. tion of hydrocortisone by ultrasonic irradiation. J Pharm Sci 56:229– 233. https://doi.org/10.1002/jps.2600560216 17 Thompson LH, Doraiswamy LK (1999) Sonochemistry: science 36 Lee J, Ashokkumar M, Kentish SE (2014) Influence of mixing and and engineering. Ind Eng Chem Res 38:1215–1249. https://doi. ultrasound frequency on antisolvent crystallisation of sodium chlo- org/10.1021/ie9804172 ride. Ultrason Sonochem 21:60–68. https://doi.org/10.1016/j. 18 Edmonds P, Dunn D (1981) Introduction: physical description of ultsonch.2013.07.005 ultrasonic fields. In: Edmonds PD (ed) Methods of experimental 37 Zeiger BW, Suslick KS (2011) Sonofragmentation of molecular physics // Ultrasonics. Acad. Press, Academic, New York, pp 1–28 crystals. J Am Chem Soc 133:14530–14533. https://doi.org/10. 19 Pohl B (2012) Charakterisierung der Mischung und Fällung bei 1021/ja205867f kontinuierlichen sonochemischen Reaktoren unter besonderer 38 Jordens J, Appermont T, Gielen B et al (2016) Sonofragmentation: Berücksichtigung der Reaktorform. Dissertation, Technische effect of ultrasound frequency and power on particle breakage. Cryst Universität Bergakademie Freiberg Growth Des 16:6167–6177. https://doi.org/10.1021/acs.cgd. 20 Apfel RE. Acoustic cavitation. In Peter D. Edmonds: Methods in 6b00088 experimental physics. Academic Press 19:355–411. https://doi.org/ 39 Jiang M, Gu C, Braatz RD (2017) Analysis of focused indirect 10.1016/S0076-695X(08)60338-5 ultrasound via high-speed spatially localized pressure sensing and 21 Kim H, Suslick K (2018) The effects of ultrasound on crystals: its consequences on nucleation. Chem Eng Process Process Intensif sonocrystallization and sonofragmentation. Crystals 8:280. https:// 117:186–194. https://doi.org/10.1016/j.cep.2017.04.008 doi.org/10.3390/cryst8070280 40 Nalajala VS, Moholkar VS (2011) Investigations in the physical 22 Brennen CE (2014) Cavitation and bubble dynamics. Cambridge mechanism of sonocrystallization. Ultrason Sonochem 18:345– University Press, Cambridge 355. https://doi.org/10.1016/j.ultsonch.2010.06.016 23 Leong T, Ashokkumar M, Kentish S (2011) The fundamentals of 41 Jordens J, Gielen B, Braeken L et al (2014) Determination of the power ultrasound - a review. Acoust Aust 39:54–63 effect of the ultrasonic frequency on the cooling crystallization of 24 Harzali H, Baillon F, Louisnard O et al (2011) Experimental study paracetamol. Chem Eng Process Process Intensif 84:38–44. https:// of sono-crystallisation of ZnSO4 7H2O, and interpretation by the doi.org/10.1016/j.cep.2014.01.006 segregation theory. Ultrason Sonochem 18:1097–1106. https://doi. 42 Jordens J, Canini E, Gielen B et al (2017) Ultrasound assisted par- org/10.1016/j.ultsonch.2011.03.007 ticle size control by continuous seed generation and batch growth. 25 Dodds J, Espitalier F, Louisnard O et al (2007) The effect of ultra- Crystals 7:195. https://doi.org/10.3390/cryst7070195 sound on crystallisation-precipitation processes: some examples and 43 Narducci O, Jones AG, Kougoulos E (2011) Continuous crystalli- a new segregation model. Part Part Syst Charact 24:18–28. https:// zation of adipic acid with ultrasound. Chem Eng Sci 66:1069–1076. doi.org/10.1002/ppsc.200601046 https://doi.org/10.1016/j.ces.2010.12.008 26 Grossier R, Louisnard O, Vargas Y (2007) Mixture segregation by 44 Li H, Wang J, Bao Y et al (2003) Rapid sonocrystallization in the an inertial cavitation bubble. Ultrason Sonochem 14:431–437. salting-out process. J Cryst Growth 247:192–198. https://doi.org/10. https://doi.org/10.1016/j.ultsonch.2006.10.010 1016/S0022-0248(02)01941-3 27 Kurotani M, Miyasaka E, Ebihara S et al (2009) Effect of ultrasonic 45 Guo Z, Zhang M, Li H et al (2005) Effect of ultrasound on anti- irradiation on the behavior of primary nucleation of amino acids in solvent crystallization process. J Cryst Growth 273:555–563. supersaturated solutions. J Cryst Growth 311:2714–2721. https:// https://doi.org/10.1016/j.jcrysgro.2004.09.049 doi.org/10.1016/j.jcrysgro.2009.03.009 46 Gracin S, Uusi-Penttilä M, Rasmuson ÅC (2005) Influence of ultra- 28 Miyasaka E, Kato Y, Hagisawa M et al (2006) Effect of ultrasonic sound on the nucleation of polymorphs of p -Aminobenzoic Acid. irradiation on the number of acetylsalicylic acid crystals produced Cryst Growth Des 5:1787–1794. https://doi.org/10.1021/cg050056a under the supersaturated condition and the ability of controlling the 47 Kordylla A, Koch S, Tumakaka F et al (2008) Towards an optimized final crystal size via primary nucleation. J Cryst Growth 289:324– crystallization with ultrasound: Effect of solvent properties and ul- 330. https://doi.org/10.1016/j.jcrysgro.2005.11.084 trasonic process parameters. J Cryst Growth 310:4177–4184. 29 Miyasaka E, Ebihara S, Hirasawa I (2006) Investigation of primary https://doi.org/10.1016/j.jcrysgro.2008.06.057 nucleation phenomena of acetylsalicylic acid crystals induced by ultrasonic irradiation—ultrasonic energy needed to activate primary 48 Louhi-Kultanen M, Karjalainen M, Rantanen J et al (2006) nucleation. J Cryst Growth 295:97–10 Crystallization of glycine with ultrasound. Int J Pharm 320:23–29. 1. https://doi.org/10.1016/j. https://doi.org/10.1016/j.ijpharm.2006.03.054 jcrysgro.2006.07.020 30 Hem SL (1967) The effect of ultrasonic vibrations on crystallization 49 Kurotani M, Hirasawa I (2008) Polymorph control of sulfamerazine processes. Ultrasonics 5:202–207. https://doi.org/10.1016/0041- by ultrasonic irradiation. J Cryst Growth 310:4576–4580. https:// 624X(67)90061-3 doi.org/10.1016/j.jcrysgro.2008.08.002 318 J Flow Chem (2021) 11:303–319 50 Hatakka H, Alatalo H, Louhi-Kultanen M et al (2010) Closed-loop flow inverter. Ind Eng Chem Res 59:3934–3942. https://doi.org/ control of reactive crystallization PART II: polymorphism control of 10.1021/acs.iecr.9b04160 L-glutamic acid by sonocrystallization and seeding. Chem Eng 70 Hohmann L, Gorny R, Klaas O et al (2016) Design of a continuous Technol 33:751–756. https://doi.org/10.1002/ceat.200900577 tubular cooling crystallizer for process development on lab-scale. 51 Nguyen T, Khan A, Bruce L et al (2017) The effect of ultrasound on Chem Eng Technol 39:1268–1280. https://doi.org/10.1002/ceat. the crystallisation of paracetamol in the presence of structurally similar impurities. Crystals 7:294. https://doi.org/10.3390/cryst7100294 71 Hohmann L, Greinert T, Mierka O et al (2018) Analysis of crystal 52 Han B, Ezeanowi NC, Koiranen TO et al (2018) Insights into design size dispersion effects in a continuous coiled tubular crystallizer: criteria for a continuous, sonicated modular tubular cooling crystal- experiments and modeling. Cryst Growth Des 18:1459–1473. lizer. Cryst Growth Des 18:7286–7295. https://doi.org/10.1021/acs. https://doi.org/10.1021/acs.cgd.7b01383 cgd.8b00700 72 Schmalenberg M, Hohmann L, Kockmann N (2018) Miniaturized 53 Amara N, Ratsimba B, Wilhelm A-M et al (2001) Crystallization of tubular cooling crystallizer with solid-liquid flow for process devel- potash alum: effect of power ultrasound. Ultrason Sonochem 8:265– opment. In: Proceedings of the ASME 16th International 270. https://doi.org/10.1016/S1350-4177(01)00087-6 Conference on Nanochannels, Microchannels, and Minichannels 54 Ruecroft G, Bruns J (2010) An apparatus and process for producing 2018. American Society of Mechanical Engineers Digital crystals (WO 2010/079350 A2). Accessed 26 Aug 2020 Collection; The American Society of Mechanical Engineers, New 55 Kudo S, Takiyama H (2012) Production of fine organic crystalline York. https://doi.org/10.1115/ICNMM2018-7660 particles by using milli segmented flow crystallizer. J Chem Eng Jpn 73 Kurt SK, Gelhausen MG, Kockmann N (2015) Axial dispersion and 45:305–309. https://doi.org/10.1252/jcej.11we168 heat transfer in a milli/microstructured coiled flow inverter for nar- 56 Hussain MN, Jordens J, John JJ et al (2019) Enhancing pharmaceu- row residence time distribution at laminar flow. Chem Eng Technol tical crystallization in a flow crystallizer with ultrasound: Anti- 38:1122–1130. https://doi.org/10.1002/ceat.201400515 solvent crystallization. Ultrason Sonochem 59:104743. https://doi. 74 Hohmann L, Schmalenberg M, Prasanna M et al (2019) Suspension org/10.1016/j.ultsonch.2019.104743 flow behavior and particle residence time distribution in helical tube 57 Savvopoulos SV, Hussain MN, Jordens J et al (2019) A mathemat- devices. Chem Eng J 360:1371–1389. https://doi.org/10.1016/j.cej. ical model of the ultrasound-assisted continuous tubular crystalliza- 2018.10.166 tion of aspirin. Cryst Growth Des 19:5111–5122. https://doi.org/10. 75 Klose A, Merkelbach S, Menschner A et al (2019) Orchestration 1021/acs.cgd.9b00466 requirements for modular process plants in chemical and pharma- 58 Rossi D, Jamshidi R, Saffari N et al (2015) Continuous-flow ceutical industries. Chem Eng Technol 42:2282–2291. https://doi. sonocrystallization in droplet-based microfluidics. Cryst Growth org/10.1002/ceat.201900298 Des 15:5519–5529. https://doi.org/10.1021/acs.cgd.5b01153 76 Elma Schmidbauer GmbH (2020) Elmasonic S 30 H: Ultraschall- 59 Koiranen T, Ekberg B, Häkkinen A et al (2018) An ultrasound Reinigungsgerät. https://www.elma-ultrasonic.com/fileadmin/ crystallization device and an ultrasound crystallization system downloads/Produktprofile/Produktprofile_DE/Ultraschallgeraete/ (WO/2018/096205). Accessed 26 Aug 2020 Elmasonic_S/PP_Elmasonic_S30H_DE.pdf. Accessed 14 Sep 2020 60 Ezeanowi N, Pajari H, Laitinen A et al (2020) Monitoring the dynam- 77 Schmalenberg M, Sallamon F, Haas C et al (2020) Temperature- ics of a continuous sonicated tubular cooling crystallizer. Cryst controlled minichannel flow-cell for non-invasive particle measure- Growth Des 20:1458–1466. https://doi.org/10.1021/acs.cgd.9b01103 ments in solid-liquid flow. In: Proceedings of the ASME 18th 61 Zierep J, Bühler K (2018) Grundzüge der Strömungslehre: International Conference on Nanochannels, Microchannels, and Grundlagen, Statik und Dynamik der Fluide, 11. Aufl. 2018. Minichannels 2020. American Society of Mechanical Engineers Springer Fachmedien Wiesbaden; Imprint: Springer Vieweg, Digital Collection; The American Society of Mechanical. https:// Wiesbaden doi.org/10.1115/ICNMM2020-1062 62 Saxena AK, Nigam KDP (1984) Coiled configuration for flow in- 78 Borchert C, Sundmacher K (2011) Crystal aggregation in a flow version and its effect on residence time distribution. AIChE J 30: tube: image-based observation. Chem Eng Technol 34:545–556. 363–368. https://doi.org/10.1002/aic.690300303 https://doi.org/10.1002/ceat.201000465 63 Klutz S, Kurt SK, Lobedann M et al (2015) Narrow residence time 79 Huo Y, Liu T, Liu H et al (2016) In-situ crystal morphology iden- distribution in tubular reactor concept for Reynolds number range of tification using imaging analysis with application to the L-glutamic 10–100. Chem Eng Res Des 95:22–33. https://doi.org/10.1016/j. acid crystallization. Chem Eng Sci 148:126–139. https://doi.org/10. cherd.2015.01.003 1016/j.ces.2016.03.039 64 Mansour M, Liu Z, Janiga G et al (2017) Numerical study of liquid- 80 Kumar V, Mridha M, Gupta AK et al (2007) Coiled flow inverter as liquid mixing in helical pipes. Chem Eng Sci 172:250–261. https:// a heat exchanger. Chem Eng Sci 62:2386–2396. https://doi.org/10. doi.org/10.1016/j.ces.2017.06.015 1016/j.ces.2007.01.032 65 Sharma L, Nigam K, Roy S (2017) Single phase mixing in coiled 81 (2013) VDI-Wärmeatlas. Springer Berlin Heidelberg, Berlin tubes and coiled flow inverters in different flow regimes. Chem Eng 82 Emig G, Klemm E (2017) Chemische Reaktionstechnik. Springer Sci 160:227–235. https://doi.org/10.1016/j.ces.2016.11.034 Berlin Heidelberg, Berlin 66 Vashisth S, Nigam KDP (2008) Liquid-phase residence time distri- 83 Gelhausen MG, Kurt SK, Kockmann N (2014) Mixing and heat bution for two-phase flow in coiled flow inverter. Ind Eng Chem transfer in helical capillary flow reactors with alternating bends. Res 47:3630–3638. https://doi.org/10.1021/ie070447h In: ASME 2014 12th International Conference on Nanochannels, 67 Rossi D, Gargiulo L, Valitov G et al (2017) Experimental charac- Microchannels, and Minichannels collocated with the ASME 2014 terization of axial dispersion in coiled flow inverters. Chem Eng Res 4th Joint US-European Fluids Engineering Division Summer Des 120:159–170. https://doi.org/10.1016/j.cherd.2017.02.011 Meeting. ASME 68 Kurt SK, Akhtar M, Nigam KDP et al (2017) Continuous reactive precipitation in a coiled flow inverter: inert particle tracking, modu- 84 Wohlgemuth K, Schembecker G (2013) Modeling induced nucle- lar design, and production of uniform CaCO 3 particles. Ind Eng ation processes during batch cooling crystallization: A sequential Chem Res 56:11320–11335. https://doi.org/10.1021/acs.iecr. parameter determination procedure. Comput Chem Eng 52:216– 7b02240 229. https://doi.org/10.1016/j.compchemeng.2012.12.001 69 Benitez-Chapa AG, Nigam KDP, Alvarez AJ (2020) Process inten- 85 Lorimer JP, Mason TJ, Fiddy K (1991) Enhancement of chemical si reactivity by power ultrasound: an alternative interpretation of the fication of continuous antisolvent crystallization using a coiled J Flow Chem (2021) 11:303–319 319 Lena K. Weick studied hot spot. Ultrasonics 29:338–343. https://doi.org/10.1016/0041- Chemical Engineering at TU 624X(91)90032-4 DortmundUniversity since 2013. 86 Valitov G, Jamshidi R, Rossi D et al (2020) Effect of acoustic stream- She graduated with the master ing on continuous flow sonocrystallization in millifluidic channels. thesis “Investigation ofdifferent Chem Eng J 379:122221. https://doi.org/10.1016/j.cej.2019.122221 operating modes of a coiled tube 87 Durga KKH, Selvarajan P, Shanthi D (2012) Nucleation kinetics, cooling crystall izer ” at XRD and SHG studies of L-Alanine single crystals grown at differ- theLaboratory of Equipment ent supersaturation levels. Int J Curr Res Rev 2012:68–77 Design in 2020. 88 Hofmann G (ed) (2004) Kristallisation in der industriellen Praxis. Wiley-VCH, Weinheim 89 Furuta M, Mukai K, Cork D et al (2016) Continuous crystallization using a sonicated tubular system for controlling particle size in an API manufacturing process. Chem Eng Process Process Intensif 102:210–218. https://doi.org/10.1016/j.cep.2016.02.002 90 Eder RJP, Radl S, Schmitt E et al (2010) Continuously seeded, continuously operated tubular crystallizer for the production of ac- tive pharmaceutical ingredients. Cryst Growth Des 10:2247–2257. https://doi.org/10.1021/cg9015788 Norbert K ockman n 91 Eder RJP, Schmitt EK, Grill J et al (2011) Seed loading effects on studiedmechanical engineering at the mean crystal size of acetylsalicylic acid in a continuous-flow the Technical University of crystallization device. Cryst Res Technol 46:227–237. https://doi. Munich and completed hisDr.- org/10.1002/crat.201000634 Ing. in process engineering at the 92 Wiedmeyer V, Anker F, Bartsch C et al (2017) Continuous crystal- University in Bremen. After 5 lization in a helically coiled flow tube: analysis of flow field, resi- years in chemicalindustry as pro- dence time behavior, and crystal growth. Ind Eng Chem Res 56: ject manager he joined Freiburg 3699–3712. https://doi.org/10.1021/acs.iecr.6b04279 University, IMTEK in 2001 93 Wiedmeyer V, Voigt A, Sundmacher K (2017) Crystal population asgroup leader for micro process growth in a continuous helically coiled flow tube crystallizer. Chem engineering. In 2007, Dr. Eng Technol 40:1584–1590. https://doi.org/10.1002/ceat.201600530 Kockmann joined LonzaLtd., Visp, Switzerland, as senior scien- tist responsible for continuous flowprocesses and microreactor Publisher’snote Springer Nature remains neutral with regard to jurisdic- technology. In April 2011, Norbert Kockmann wasappointed as full pro- tional claims in published maps and institutional affiliations. fessor for equipment design at TU Dortmund University,Germany. His research interests are on small-scale device for continuouschemical pro- cesses, modular design, and process intensification. His workincludes Mira Schmalenberg studied fundamental investigations of small-scale multiphase flow, modellingand Chemical Engineering at TU simulation accompanied by modern sensing technology and machine- DortmundUniversity. In November learningmethods. 2017, she completed her master the- sis on the design ofa modular con- tinuous-flow tubular cooling crys- tallizer for process developmentand small-scale production at the Laboratory of Equipment Design. In February2018, she started as a research assistant in the ENPRO2.0-TeiA(‘‘Trennverfahren mit effizienten und intelligenten Apparat en ’’ ) project at theLaboratory of Equipment Design. Her focus is theinvestigation of small scale continuous cooling crys- tallizers. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Flow Chemistry Springer Journals

Nucleation in continuous flow cooling sonocrystallization for coiled capillary crystallizers

Loading next page...
 
/lp/springer-journals/nucleation-in-continuous-flow-cooling-sonocrystallization-for-coiled-eeEgqTrOj1
Publisher
Springer Journals
Copyright
Copyright © The Author(s) 2021
ISSN
2062-249X
eISSN
2063-0212
DOI
10.1007/s41981-020-00138-x
Publisher site
See Article on Publisher Site

Abstract

Nucleation in continuously operated capillary coiled cooling crystallizers is experimentally investigated under the influence of ultrasound. It was found that there is no sharp boundary but rather a transition zone for nucleation under sonication. For this purpose, a tube with an inner diameter of 1.6 mm and a length of 6 m was winded in a coiled flow inverter (CFI) design and immersed into a cooled ultrasonic bath (37 kHz). The CFI design was chosen for improved radial mixing and narrow residence time distribution, which is also investigated. Amino acid L-alanine dissolved in deionized water is employed in a supersaturation range of 1.10 to 1.46 under quiet and sonicated conditions. Nucleation is non-invasive detected using a flow cell equipped with a microscope and camera. . . . . . Keywords Coiled flow inverter Continuous nucleation Cooling crystallization Minichannel Capillary flow Non-invasive nucleation detection Sonocrystallization Introduction For continuous operation in crystallization, many aspects have to be considered. Clogging should be avoided, while a The continuous operation of small-scale equipment is of ut- narrow crystal size distribution (CSD) is targeted to simplify most importance for accelerated process development [1–4]. It and minimize post-crystallization operations [7, 8]. In addi- is not necessary to switch the operation mode from small scale tion, high yield, high purity, and specific particle morphology batch to the continuous production of bigger quantities, are often required for sufficient product quality. To avoid the smallest quantities can already be continuously produced, preparation of seed crystals some concepts have already been and similar equipment concepts can be transferred more easily developed to integrate the nucleation into the crystallization to larger scales [5]. Furthermore, the continuous mode has process [9–14]. advantages, such as smaller hold-ups, higher product repro- The concept of continuous nucleation by sonication at con- ducibility, and space-time-yield [1, 2, 5, 6]. tinuous flow with good mixing caused by the coiled flow inverter (CFI) design is evaluated in more detail in this con- tribution. The main goal here is to design an apparatus concept Highlights that allows for the non-invasive observation of nucleation due � Sonicated continuous flow for nucleation in coiled flow inverter. to sonication. In the long term, this forms a database which � Nuclei detection non-invasive in continuous flow. will enable a suitable operation for the production of continu- � Determination of two primary nucleation thresholds under sonication, one at the lower limit (transition region starts) and one where nucleation ous seed crystals. occurs in every experiment under given conditions. Mira Schmalenberg and Lena K. Weick contributed equally to this work. Mechanisms of sonocrystallization in solution * Mira Schmalenberg The application of ultrasound for process intensification is a mira.schmalenberg@tu-dortmund.de frequently used and increasingly popular tool [15, 16]. * Norbert Kockmann Ultrasound is defined as sound waves in a frequency range norbert.kockmann@tu-dortmund.de of 16 kHz to 500 MHz [17]. For the transmission of these waves a medium, for example water, is always required BCI Equipment Design, TU Dortmund University, Emil-Figge-Straße 68, 44227 Dortmund, Germany [18]. It is assumed that two phenomena occur simultaneously 304 J Flow Chem (2021) 11:303–319 but can vary strongly depending on the conditions. On the one fragmentations occur for transient and stable cavitation [38]. hand, ultrasound can cause increased molecular movement in With the distinction between transient and stable cavitation, the fluid, which can lead to an increase in the mass transfer as the transient cavitation was observed at 30 kHz and from well as increased mixing [19]. On the other hand, acoustic 41 kHz to 1.14 MHz there is a rather stable cavitation [16]. cavitation bubbles are generated by the impact of ultrasound The collapse intensity of cavitation as well as the active bub- [20]. A detailed description of cavitation bubble formation is ble size and its distribution depend on the ultrasound frequen- given in [21–23]. These cavitation bubbles collapse almost cy, the energy, and the solution itself [16]. Additionally, it has − 1 immediately after they are formed [16, 23], which leads to already been found as rule-of-thumb 35 W∙L (for most sol- local maxima in temperature, while pressure hotspots occur vents operating at room temperature) [15] must be present for during this process [21]. Several theories describe the mecha- cavitation bubbles to form [17, 19]. nism of nucleation by cavitation, but “there is still no consen- sus on the fundamental mechanism of ultrasound action on the Influence parameters of sonocrystallization crystallization process” [16]. The possible processes that are probably influenced by ultrasound during the crystallization Parameters that can be varied and adjusted from the external process are briefly described here and more detail on the pro- environment are the ultrasonic hardware, the frequency, the cesses can be found in [16]. ultrasound exposure, the energy input, and the supersaturation. Concerning the primary homogeneous nucleation, it is dif- Depending on these parameters the induction time, the meta ficult to observe it because there can always be the smallest stable zone width (MZW), the nucleation rate, the polymor- impurities in the system, which would mean heterogonous phism, the yield, the crystal size, and its distribution can be nucleation was detected. A presumed mechanism for homog- influenced. The different ultrasonic system types are ultrasonic enous nucleation is the “segregation theory”, where it is as- baths, which were originally manufactured for equipment sumed that crystal clusters are formed under the effect of an cleaning, probe (horn) systems, and planar transducers [17, acoustic pressure gradient until they result in a larger crystal 21]. The most important advantage of the ultrasonic horn and [16, 24–26]. Other studies show that the acoustic irradiation plate is that the ultrasound can be induced more effectively and leads to overcoming of the critical growth energy [16, 27–29]. with a better focus [12, 39]. The main advantage of the ultra- Most authors postulate that heterogeneous primary nucleation sonic bath is that it is standard laboratory equipment, which is by sonication is present instead of homogeneous primary nu- often already available and has not to be purchased separately. cleation. It is assumed that the reactor surface and smallest The frequency is an important setting for sonocrystallization impurities can reduce the surface energy barrier [16, 30]. because it can affect the bubble dynamics [21, 40], which can Additionally, dissolved gases in the liquid could act as cavi- affect the crystallization, too. The MZW could be reduced in tation nuclei [19]. Therefore, degassing of the solution or fil- comparison to silent conditions for different frequencies (range tering can reduce cavitation bubble formation [16, 31]. of 41–1140 kHz) with constant calorimetric power [41]. The Inversely, investigations on gassing crystallization have higher the used frequency, the less the MZW could be reduced shown that active gassing of a solution can lead to results [41]. If the applied frequency is decreased the formed crystals similar to those of sonocrystallization [32, 33]. become smaller [42]. Nevertheless, Kim and Suslick point out The investigations on cavitation bubble formation using that “the number of cavitating bubbles is not controllable, and sonochemiluminescence by Sarac et al. [34]showed bubble so it is exceedingly difficult to compare different frequencies formation solely using free interfaces generated by slug-flow. due to changes in the number of cavitating bubbles, which are A coiled perfluoroalkoxy alkane (PFA) tube with an inner highly dependent on the specific apparatus used” [21]. Besides diameter of 0.8 mm was placed in an ultrasonic bath the frequency, the intensity is an often-used setting for (27.2 kHz) [34]. No cavitation bubble formation was observed sonocrystallization investigations. In general, it was determined without the introduction of a gaseous phase [34]. that crystals become smaller with increasing US intensity [21]. Regarding secondary nucleation, it is assumed that ultra- Since in this work an ultrasonic bath was used that has sound increases erosion, fracture, and abrasion, resulting in frequency and intensity pre-set, the settings which are more the formation of new nucleation clusters [16, 35]. This in- independently adjustable will be discussed. This includes for crease is due to better mixing by ultrasound, which inter alia example the acoustic irradiation time. It was found that with increases the possibility of crystal collision. However, the increasing sonication time the crystal sizes decrease and be- same mixing effects produced with ultrasound may not pro- come more uniform [21, 43, 44]. duce the same results as mechanical mixing [16, 30, 36, 37]. The energy input is another important setting that should During crystal growth and agglomeration or fracture, the always be taken into account when considering mechanisms “sonofragmentation, sono-deagglomeration and sonocrystallization [17]. In most cases, this parameter results sonoagglomeration” can beobserved, a moredetaileddescrip- from the design of equipment and the hardware used including ultrasound intensity and frequency. However, it can also be set tion can be found in Nalesso et al. [16]. Different types of J Flow Chem (2021) 11:303–319 305 constant to explicitly exclude this influence [41]. Miyasaka was patented, which is equipped with an ultrasonic transducer et al. and Kurotani et al. explicitly investigated the influence and designed with a continuous flow, mainly used for nucle- of the input calorimetric energy in connection with primary ation [54]. nucleation [27, 29]. The calorimetric energy P is determined Narducci et al. were among the first to design a continu- cal − 1 according to Eq. (1) by recording the temperature T change in ously (10–30 mL min ) operated and with 20 kHz ultrasonic the medium as a function of time t and by multiplying the heat processor sonicated stirred vessel (300 mL). In this study, the capacity c of the fluid and the heated mass m. influence of the duration of exposure, the achievement of the steady-state, the crystal habitus, and the particle size distribu- dT tion were already examined. They found out, that the particle P ¼ c m ð1Þ cal p dt size under sonication resulted in smaller crystals [43]. Kudo and Takiyama [55] investigated the generation of For two amino acids, for example, it was found that there is a crystal nuclei for anti-solvent crystallization in segmented local maximum for induction time depending on the ultrasound flow in a metallic and a flexible tube (last one for connection), energy, which in turn is dependent on the degree of supersatura- each with an inner diameter of 2.4 mm. The metallic tube (L = tion and the model system used [27]. The supersaturation itself is 0.5 m) was coiled into a 42 kHz ultrasonic bath [55]. The main also an externally adjustable influencing setting. Jordens et al. results were that monodispersed crystal size distribution showed that the greater the supersaturation, the smaller the crys- (CSD) was created [55]. Further investigations on ultrasonic tals become, whereby an asymptote was detected from supersat- anti-solvent crystallization in milli structured equipment can urationof1.56[42]. Target parameters that are influenced by be found in [56, 57] from the van Gerven group. ultrasound are induction time, the MZW, the nucleation rate, At about the same time the research group of Gruber- the polymorphism, yield, particle size, and its distribution along Woelfler worked on continuous cooling crystallization of with other interactions generated by ultrasonic irradiation [16]. acetylsalicylic acid in ethanol using slug-flow and sonication Ultrasonic irradiation affects the induction time not only − 1 with a flow rate of 6 mL min . The coiled polysiloxane tube for anti-solvent crystallization [44, 45] but also for cooling with an inner diameter of 2 mm and with 1.5 m length was crystallization [27, 46] and usually shortens the induction time inserted into a 35 kHz ultrasonic bath. One point of investiga- probably due to the better micro-scale mixing and the turbu- tion was the nearly constant CSD over the required time lent flow generated by the acoustic cavitation [21]. (25 min). The segmented flow was implemented after the The MZW also decreases under the influence of ultrasound nucleation section and it followed a coiled tubular crystallizer [41, 46, 47]. It was found that higher reduction of MZW was at four different temperature stages. For the whole crystalliza- achieved at low frequencies (41 kHz) [38]. The nucleation rate tion process, different seed loadings were investigated with is presumably increased because the critical excess free energy their influence of a fine dissolution after the ultrasonic bath. is reduced, as the molecule is only surrounded by solution on Furthermore, they analyzed the effect of different ratios of one side and is in contact with the cavitation bubble on the seed-mass and the effect of ultrasound on the product crystals other [21]. Secondary nucleation is also increased because itself. The authors could show that it is possible to design a mixing is improved and the number of collisions is raised [21]. continuously seeded flow-through crystallizer with in-situ So far unknown is the effect on polymorphic forms by seed generation from ultrasound resulting in a constant seed sonication, it is just determined that ultrasound can affect the quantity and quality [11]. polymorphic form [46, 48–50]. A further product-determining A subsequent study, also by Khinast’s research group, factor is the yield, which can be higher with acoustic irradia- investigated the controllability of crystal size in the con- tion because ultrasound reduces induction time so nucleation tinuous tubular crystallizer. They used the same tubes begins earlier [51, 52]. The particle size and its distribution is a (d = 2 mm, polysiloxane) for the coiled crystallizer and very important quantity, also for further processing, so this is a i the same cooling crystallization model system widely studied parameter [27, 40, 42, 43, 51, 53]. Besides, (acetylsalicylic acid, ethanol). However, the seed genera- other interactions can be influenced by ultrasound for example tion via ultrasound was done in a round-bottom flask in interparticle collision, direct interaction between particles, and 500 mL, which was embedded into the ultrasonic bath shockwaves [21, 37]. (35 kHz). Here the supersaturation u was given and the ultrasonic irradiation duration was 60–70 s. The feed so- − 1 Continuous sonocrystallization in capillaries lution flow rate was 22 mL min and the seed suspen- − 1 sion flow rate was 6 mL min , these both flows were Since there are already some studies on nucleation by the mixed before entering the crystallization section (three influence of ultrasound irradiation, the most important ones, thermostat baths each with 5 m coiled tube). At the outlet, considering the setup featured in this work, are presented here. the CSD was evaluated and the crystal size could be con- Already in 2010, an Ultrasonic Oscillatory Baffled Reactor trolled by varying the volume flow rate of the seeds. To 306 J Flow Chem (2021) 11:303–319 avoid clogging, the crystallizer was rinsed with solvent choice and description of the winding type, which will be approximately every 10 minutes [13]. described in more detail below, this and the associated flow Nucleation experiments for different supersaturations in a conditions are simply repeatable. The non-invasive and in-situ range from 1.12 to 1.81 were carried out in another milliscale analysis of the crystals by image analysis ensures that really apparatus, a glass flow cell connected to a PFA tube (d = only crystals are produced in the apparatus and not in the 5 mm), and a stirred vessel, here for the model system para- periphery. The investigations also close the gap that, until cetamol in water. The ultrasonic source was an ultrasound now, crystallization investigations have only been studied transducer (20 kHz), which was mounted on the glass flow for selected supersaturations and not for a merging range. cell or on the stirring tank. The flow rate was selected to be This shows that due to the stochastic process of nucleation, − 1 100 mL min . With ultrasound, nuclei could be generated there is no clearly defined boundary between no crystal for- already from supersaturation of 1.32, and under silent condi- mation and crystal formation. The choice for the continuous tion a supersaturation of 1.87 was necessary [42]. mode of the experiments is not only the basis to construct a Rossi et al. [58] investigated non-invasive with a high- seed crystal unit, but also takes advantage of the fact that many speed microscope cooling sonocrystallization of adipic acid individual batch experiments can be replaced by less contin- in water in diverse supersaturations with a sonoprobe uous runs. (20 kHz). They used a straight process tube made of PFA with an inner diameter of 1 mm and investigated liquid-liquid slug- flow with n-hexane as an inert carrier fluid with a flow rate of Fluid dynamics in a coiled flow inverter − 1 0.3925 mL min [58]. Another research group has recently released a patent, it Many of the discussed research projects used a coiled pipe or comprises the inlet of a substance to be crystallized into a tube for the process flow. It is space-saving and the flow itself tubular crystallizer which is sonicated (20 to 40 kHz) by an exhibits good radial mixing. In capillaries, the velocities are ultrasonic source and includes a temperature control structure usually low with laminar flow, hence, in the macroscopic for cooling the process fluid [59]. Studies on this patent were view, there is low radial mixing in the flow of straight capil- made by Han et al. and Ezeanowi et al. [52, 60]. laries [61]. Han et al. use three ultrasonic baths (35 kHz) connected in To create a macroscopic exchange in the flow, coiled tubes series and equipped with thermostats. In each ultrasonic bath, are used, especially coiled flow inverters (CFI). They are con- a 12 m or 18 m coiled stainless-steel tube with an inner diam- structed of helical segments with a minimum of three turns eter of 4 mm was inserted. Three different substance systems and the segments are linked by 90° bends. The coiled structure namely potassium sulfate, copper sulfate, and phthalic acid itself generates a centrifugal force on the flow, which causes dissolved in water were examined for cooling crystallization. radial mixing, the so-called Dean vortices. Due to the 90° − 1 The mass flow rate was 75 mL min . The residence time bends the direction of the Dean vortices is changing so the dependence, the cooling strategy, the influence of supersatu- radial mixing is improved [62]. ration (1.2, 1.36, 1.54, 1.71, 1.77, and 1.9), and the acoustic Many investigations about coiled tubes and CFIs have been irradiation itself were investigated. No usable crystals could made, especially in characterizing the improved mixing and be produced without sonication. Additionally, no nuclei were the residence time, which could be narrowed by using a CFI formed at a supersaturation of 1.2, and the system was blocked instead of a straight tube [62–67]. Some studies have used the at a supersaturation of 1.9 [52]. CFI design for precipitation [68, 69]orcrystallization Ezanowi et al. investigated a bigger version of the setup in [70–72]. pilot scale with an inner diameter twice as large and approx- The main key numbers used to describe a CFI are briefly imately equal length. They found out that with increasing presented here. The Dean number Dn describes the intensity residence time the CSD decreased and that ultrasound of the Dean vortices formation (Eq. (2)). It is defined as prevented pipe clogging [60]. Reynolds number Re multiplied with the square root of the In most of the studies, few information was provided on ratio of the inner diameter d to the coil diameter d . i c how exactly the tube was coiled. In addition, the analysis was sffiffiffiffiffi usually carried out offline, for example by collecting the prod- d Dn ¼ Re  ð2Þ uct solution in a filter or by using a temperature-controlled collecting-vessel, which both does not display the in-situ mo- The Reynolds number is known as the ratio of inertia force ment. In all the studies described, the analyses have been carried out for selected supersaturations, but it has not been to frictional force (Eq. (3)) and described as mean velocity u multiplied with liquid density ρ and characteristic length, here investigated whether there is a transition or a precise boundary from which nuclei are generated by ultrasound. To fill the the inner diameter, divided by the dynamic viscosity of the liquid η [61]. gaps addressed here, this work was performed. Under exact l J Flow Chem (2021) 11:303–319 307 uρ d l The CFI Design was chosen for its good mixing properties Re ¼ ð3Þ η and the low space consumption caused by the coiled tube. The experimental design is described below. The small inner di- To ensure a narrow residence time distribution in a CFI, ameter (d = 1.6 mm) for the process tube was selected due to − 1 various requirements must be fulfilled. The Dean number has theneed for asufficiently high flow rate (around 16 g min ) to be equal to or higher than three [73], the number of turns to guarantee a homogeneous suspension flow (for the chosen has to be equal to or more than five [73], and the number of model system and 1 w.% solid). Besides, preliminary tests for bends has to be equal to or more than three [62]. Additionally, this combination have already been carried out on a cooling the modified torsion number T , see Eq. 4, has to be equal or crystallizer [72], which in the future will also be connected to higher than 1000, where t is the helix pitch [63, 68]. the plant presented here. After the description of the experimental setup, the analyt- T ¼ Re  ð4Þ ical methods used are presented and the performance regard- ing residence time measurements, nucleation experiments, and calorimetric power is described. In this article, water al- If solids are present in the flow, it is important to avoid ways means deionized water. clogging, a homogeneous suspension flow must be ensured, and no particles should sediment. With an empirical correla- Experimental set‐up tion based on the tube inner diameter, the mean particle size, the particle mass fraction, the suspension flow rate, and the The structure of the plant is divided into three units (Fig. 1). pulsation characteristics (using a peristaltic pump), a homoge- neous suspension flow could be proclaimed. To ensure homo- The first unit consists of the feed and pump unit (FPU), the second is the ultrasonic unit (USU) and the third unit is the geneous suspension flow, the critical Froude number Fr , d,crit product analysis unit (PAU). In keeping with the modular which was developed as an empirical correlation, can be de- concept, this structure would consist of three Process termined to find out, which maximum mean particle size can Equipment Assemblies (PEAs), which could be used as a be transported homogeneously in the tube with a given solid mass fraction. In Eq. (5), the empirical correlation for the single module [75]. The FPU consists of two double-jacketed glass storage critical Froude number is given with model parameters for the switch from moving sediment flow to homogeneous sus- tanks with a filling volume of either 1 or 3.5 L. Here, the feed solution and the solvent are prepared, for which they are pension flow [74]. placed on magnetic stirring plates (MR 3001, Heidolph –4:564 0:717 Instruments GmbH und Co. KG, Germany). The storage tanks Fr ¼ 0:252  Re ðÞ 1–w ð5Þ d;crit s each contain a triangular magnetic stirring rod (PTFE, 50 mm, With the usual definition of the densimetric Froude number Bohlender GmbH, Germany). The tanks are temperature con- trolled with a thermostat (CC304, Huber Kältemaschinenbau (Eq. (6)), the maximum mean particle size x can be deter- mined by converting Eq. (6). The necessary variables are the AG, Germany) and they are connected in series (first the feed storage, second the solution storage). The feed tank contains Reynolds number, the used solid weight fraction w , the mean velocity of the process flow, the liquid density, the density the undersaturated L-alanine water solution, while the solution difference between the solid and the liquid phase Δρ and tank is filled with water for temperature equilibration and sl the acceleration of gravity g [74]. cleaning. The solution is pumped via a peristaltic pump (LabDos Easy-Load, HiTec Zang GmbH, Germany), using a Tygon® tube (inner diameter d 1.6 mm). To ensure that the u ρ i Fr ¼  ð6Þ solution does not contain particles when it enters the process x g Δρ 50;3 sl tube, it is preceded by a filter frit (pore size III, ROBU Experimental section and methods Glasfilter Geräte GmbH, Germany). The process tube is made of fluorinated ethylene propylene (FEP) (Bohlender GmbH, This contribution presents an experimental setup of a coiled Germany) and has an inner diameter d of 1.6 mm. The FPU is tube within an ultrasound bath with an adjacent measuring cell connected with a three-way connector for a pressure sensor below a microscope, which allows in-situ and non-invasive integrated to the third link (A-10, WIKA, Germany) to record observation of the nucleation. To evaluate the flow conditions the pressure loss. The tube from the feed storage to the three- in the coiled tube, the residence time behavior is analyzed. way connector has a length of 1.6 m. Nucleation experiments depending on supersaturation inves- The process tube is coiled in CFI design on a grid basket, tigated whether there is a nucleation threshold or a transition which is available for the used ultrasonic bath. This can be region. Also, particle size distribution, pressure loss, and the seen in Fig. 2 in the upper picture. The total length of the tube time till clogging appears are examined and evaluated. is 6.01 m, including 0.19 m for the inlet and 0.22 m for the 308 J Flow Chem (2021) 11:303–319 Fig. 1 Experimental Set-Up with three units, feed and pump unit (FPU), ultrasonic unit (USU) and product analysis unit (PAU), process flow from left to right outlet. The inlet and outlet of the process tube are insulated to determine the process mass flow rate from the mass with polyethylene foam. The CFI winding on the grid is di- difference with time. vided into six segments with five turns each. The coil diameter Three resistance thermometers (Pt-B-100-2. RÖSSEL- d is 52 mm and the distance between the coils, the pitch, is Messtechnik GmbH, Germany) are used to monitor the temper- about 3.6 mm. The coiled tube on the grid basket is embedded ature with one in the feed or solution storage, one in the ultra- in the ultrasonic bath (Elmasonic S30H, Elma Schmidbauer sonic bath, and one in the direct environment. Before the ex- GmbH, Germany) filled with deionized water. The bath has a perimental runs, all temperature sensors were calibrated by a filling volume of 2.75 L [76], whereas it is only filled with two-point calibration procedure with two thermostats. The inlet 2.19 L so that the coiled tube is always fully covered with and outlet temperature of the process medium are recorded via water. Moreover, the tube does not touch the ground or walls sealed thermocouples (Typ K, 1 mm diameter, OMEGA of the ultrasonic bath. According to the manufacturer, the Engineering, United States). From preliminary tests, the follow- ultrasonic bath has a frequency of 37 kHz, the total power ing correction Eq. 7 of the measured temperatures on the tube, consumption is 280 W, while 80 W being the effective ultra- and the actual temperature could be determined. This correction sonic power [76]. To ensure a constant temperature in the is valid in the temperature range 10 °C to 60 °C. ultrasonic bath for the nucleation experiments, it is connected to a cryostat (Pilot ONE ministat 125, Huber Kältemaschinenbau T ½ °C ¼ 1:1774  T ½ °C −3:9049 ð7Þ fluid surface AG, Germany). The internal pump rate is set to 3000 rpm. For the process control system, a LabManager® (HiTec The process tube, which comes out of the USU, is con- Zang GmbH, Germany) is used, data evaluation is enabled nected to a FEP tube (same inner diameter) of 0.76 m in LabVision® (HiTec Zang GmbH, Germany). The data is length and leads to the PAU. The next part of the tube recorded every second. (1.2 m) leads from the measuring cell (Fig. 2 bottom) into the sampling tank. The section of the tubing located inside the measuring flow cell has a length of 325 mm. The mea- Measurements suring flow cell in combination with a microscope (Bresser Science ADL 601P, Bresser GmbH, Germany), a photo Pressure loss camera (Nikon Z6, Nikon GmbH, Japan) and an image evaluation tool [77] based on the algorithms of Borchert In addition to temperature monitoring, the pressure loss Δp and Sundmacher[78]and Huoetal. [79] enables a non- for the experiments is measured by the pressure sensor (A-10, invasive in-line measurement of the crystals produced. WIKA, Germany). The pressure loss can also be correlated to For further details refer to [77]. The processed fluid is the mean flow velocity u according to Eq. (8), where f is the finally collected in a collecting container. This is located loss coefficient, ρ the fluid density, and L the length of the on a scale (Kern 572, KERN & SOHN GmbH, Germany) tube[61]. J Flow Chem (2021) 11:303–319 309 Fig. 2 Process tube coiled on a grid basket in CFI design (upper), measurement cell with process tube under the microscope (bottom) special temperature control of the measurement cell was nec- ρL essary although this is possible [77]. Δp ¼ f u ð8Þ 2d The photos taken are evaluated using a routine, which is based on the algorithms by [78]and [79]. This semi-automatic A suitable correlation for the friction factor f for coiled image processing routine determines the circular area equiva- tubes is Eq. (9)[80] and has already been used successfully lent diameter, using the projection surface of the crystals [77]. in other studies with tubes coiled in CFI design [63, 72]. It is composed of the pipe friction coefficient for straight pipes f , with 64/Re for laminar flow, and the Dn number. Concentration determination via gravimetric method 0:603 f ¼ f 1 þ 0:0456  Dn ð9Þ a 0 A gravimetric method is used to determine and check the Another correlation for the friction factor f for coiled tubes b dissolved amount of the model substance (L-alanine). and laminar flow is Eq. (10)[81]. Therefore, the quotient of the mass of the target component i and the mass of the solution sol is determined by Eq. (11). f ¼ f 1 þ 0:033 ðlog ðDnÞÞ ð10Þ b 0 10 m m  m i dried empty w ¼ ¼ ð11Þ Both correlations are used and compared with each other in m m  m sol sample empty order to enable better predictions to be made for future operations. An empty (and labeled) 1.5 mL reaction vessel is weighed with a scale (XA 205 Dual Range, Mettler Toldeo, US) to determine m .To get m a syringe with pre-filter (cel- empty sample Particle measurement via image analysis lulose acetate membrane, pore size 0.2 µm, VWR International, US) is filled with the solution sample (about To investigate the formed crystals, an image analysis protocol, 1 mL) and injected into a reaction vessel, and weighed again. which was already mentioned in the section above, was used. After drying the sample in a vacuum drying oven (Memmert The special property here is that the crystals can be detected VO400, Memmert GmbH + Co. KG, Germany) at 60 °C and non-invasive without large equipment effort. As a major com- with stepwise reduction of the pressure (1 bar, 600 mbar, ponent, a microscope is equipped with a camera, and a mea- 300 mbar), it will be weighed again to get m .Each sample dried surement cell is placed below. The process tube is clamped is dried for a minimum of 14 days and is considered un- through the measuring cell, which is out of glass and filled changed if the sample does not change for more than with water to observe the crystals distortion-free. Because the 0.001 g, almost the same procedure has already been success- system was mostly cooled down to room temperature, no fully carried out in [74]. 310 J Flow Chem (2021) 11:303–319 Experimental procedures method. This number (Eq. 17) describes the relationship between convective mass transport and axial mixing, First of all, the experimental procedure to determine the which includes the axial dispersion coefficient. It is also residence time of the liquid phase in the USU is described, used to indicate if axial mixing is negligible compared to followed by explaining the performance of nucleation exper- convection, i.e. the residence time behavior is comparable iments with and without ultrasound. The most important with an ideal plug flow (Bo > 100) [74, 82]. parameters are also defined here. Finally, the procedure is presented for determining the dissipated energy by the ultra- t ¼ ∫ ðÞ 1−FtðÞ dt ð14Þ t¼0 sonic bath. θ ¼ ð15Þ Residence time behavior pffiffiffiffiffiffi 1 Bo 1−θ pffiffiffiffi FðÞ θ ¼ 1–erf  ð16Þ 2 2 θ To determine the actual liquid residence time t and the distri- bution sum function F(t), the step-response method was used. uL For this purpose, deionized water (from solution storage) is Bo ¼ ð17Þ ax first introduced into the entire apparatus (USU). The inlet in the FPU was already filled with tracer solution (from feed For the description of the residence time distribution in a storage) up to the second three-way valve (see Fig. 1)in order CFI, it could already be shown that the dispersion model is to measure only the residence time distribution in the USU well suited [72, 74, 83]. (pressure sensor was not used here). As a tracer, a thiosulfate − 1 solution with a concentration of 0.2–0.4 mg·g water (not changing the parameters of water significantly) [73]is used Nucleation experiments and detected by a UV-Vis detector (Evolution 201, Thermo Fisher Scientific Inc., US). Instead of the microscope in the An amino acid (L-alanine, purity ≥ 99.6%, Evonik Rexim PAU, the UV-Vis detector with a measuring flow-through (Nanning) Pharmaceutical Co., Ltd., China) and water were cuvette (170700-0.2-40, Hellma Analytics, Germany) is used. used for the nucleation experiments with and without ultra- − 1 The mass flow rate is set to 16.5 ± 0.8 g min .To inves- sound. Based on the empirically determined solubility curve tigate the influence of the US on the residence time the exper- (Eq. (18)) according to Wohlgemuth et al. [84], the L-alanine/ iments were done three times without ultrasound and four water solution was prepared for the desired saturation temper- times with ultrasound. The ultrasonic bath was cooled to room ature T . temperature with the connected cryostat. −3 * For calculating the hydrodynamic residence time τ * ðÞ 9:0849210 T ½ °C w½ g  g ¼ 0:112381  e ð18Þ sol (Eq. (12)), the known geometry of the tube (d = 1.6 mm, L = 6 m) for the internal volume V was considered as well as the The preparation of the solution in the FPU was done one liquid density ρ , which is almost identical to that of water. day before the experiments started. Therefore, the feed tank The values for water are taken from [81]. temperature (TI_00) was set to 5 K higher than the chosen Vρ l saturation temperature (T = 50°C or60°C).Thisensures τ ¼ ð12Þ m that the crystals are all dissolved and that no nuclei were al- ready present in the feed tank. Hence, the solution is not sig- The distribution sum curve is defined as Eq. 13, with con- nificantly oversaturated before entering the ultrasonic bath. centration w at time t in dependence of the tracer concentration The temperature of the ultrasonic bath was set for each exper- w . iment, depending on the target supersaturation, for example to 20 °C (ΔT = 30 K) for a supersaturation of 1.3. To determine wtðÞ FtðÞ ¼ ð13Þ the supersaturation from temperature S , Eq. (19) was chosen, using the assumption that the activity of the dissolved compo- The actual residence time t (Eq. 14) was determined by nent behaves ideally [84]. Since this is only a value calculated integration of the step-response, using the trapezoidal from the measured temperatures, the gravimetric supersatura- rule. The use of the dimensionless time θ (Eq. 15)isused tion was determined additionally. Therefore, instead of the to compare each experiment. With the use of the disper- calculated solved fraction of L-alanine (using Eq. 18), the sion model Eq. 16, which allows to consider the axial gravimetrical measured fraction of the feed tank w was used. back mixing, the Bodenstein number Bo was determined The determined supersaturation is labeled as S in the follow- by fitting the experimental data using the least square ing (Eq. (20)). J Flow Chem (2021) 11:303–319 311 * * w T could be easily detected whether nucleation has occurred in S ¼ ð19Þ wTðÞ the case of different supersaturations. out S ¼ ð20Þ G Residence time behavior wðÞ T out The residence time distribution of the CFI in the USU is in- The sonicated and silent nucleation experiments are done vestigated for two reasons. On the one hand, the apparatus is in a supersaturation range of 1.10 to 1.36 for a saturation to be characterized entirely and the hypothesis of a narrow temperature of 50 °C and some extra investigations are made residence time will be confirmed by the CFI design. On the with an inlet saturation temperature of 60 °C for the supersat- other hand, the observations regarding acoustic streaming, uration 1.38, 1.42, and 1.46. The supersaturation was adjusted mentioned below, made by Valitov et al. [86] will be verified. via the end temperature. Before each experiment, the temper- The residence time behavior of the tube in CFI design is ature profile runs in with water until it was stationary (for shown in Fig. 3. Here the residence time sum curve under minimum of 20 min) before switching to the undersaturated quiet and sonicated conditions is plotted as a function of di- solution and starting the nucleation experiments, where each mensionless time, each with the experimental data (symbols) experiment is either run till clogging or 20 min if no crystals and fitted to the dispersion model (green line). It can be seen were visible. that the residence time distribution curve is very narrow and similar to an ideal tube. No optical differences between the Calorimetric energy input silent and sonicated curve can be seen. The most important parameters are listed in Table 1. As already mentioned, Eq. (1) is used to determine the energy − 1 The mass flow rate of 16.5 ± 0.8 g min could be kept input by the ultrasonic bath. This method assumes that the almost constant during the different tests. This allows a good whole power entering the solution is dissipated as heat [42, comparison between the silent and sonicated tests. Using the 85]. Therefore, two setups were used. Case 1 only includes the mass flow rate, the Reynolds number, Dean number, and USU and the 2nd case is the same setup as described in the modified torsion number can be determined to 238, 42, and experimental setup section. In both cases, the ultrasonic bath 10,785 respectively. These values show that the flow is in the itself was not temperature controlled via the cryostat. laminar range and the CFI design fulfills the requirements for The environmental temperature, the ultrasonic bath temper- a narrow residence time distribution. ature, and in the second case also the in- and outlet tempera- The hydrodynamic residence time is slightly shorter than ture of the process flow were measured. In every experiment, the real evaluated residence times. This could indicate that the temperature change was recorded in dependency of the there might be a small death zone; probably it is the flow- time. For every experiment, first, the ultrasonic bath tempera- through cuvette in the UV-Vis detector itself. The actual res- ture is prepared beneath or at room temperature. Then, the idence times for silent and sonicated conditions are also very ultrasonic bath was turned on. For the second case, a process similar, both within the error tolerance, but the sonicated one flow (water) with room temperature through the CFI tube is slightly smaller. The calculated Bodenstein numbers are − 1 (16.7 ± 0.1 g min ) is turned on. Each of these cases were higher than 100 for both cases, hence, the residence time be- measured four times and for a minimum time of 20 minutes. havior is similar to an ideal plug flow. Additionally, the one The performance of the ultrasonic bath was recorded for for the sonicated conditions is a bit higher than for the silent, comparability with other studies. The measurements showed a but in the range of the standard deviation, too. Therefore, it − 1 temperature change of 0.007 K·s . According to Eq. 1,this can be concluded, that the sonication has no significant effect results in an energy of 0.35 W applied to the volume of the on the residence time distribution if the CFI design is chosen. tube in the ultrasonic bath using the heat capacity of water Valitov et al. [86] investigated and evaluated the influence [81]. of the ultrasound (horn, 20 kHz) on the flow of a straight FEP tubing with inner diameters of 1.55 mm and 3.2 mm for a sonication time of 0.3, 1.0, and 3.0 s. They assumed that Results and discussion boundary layer streaming could be neglected because FEP as a material is acoustically transparent and boundary layer To verify a narrow residence time distribution in the CFI the streaming in millifluidic devices tends to be visible only at residence time behavior was investigated. Furthermore, it was higher frequencies (> 1 MHz). Furthermore, they considered evaluated whether it is possible to investigate nucleation in the and evaluated the Eckart streaming and microjetting, but for described setup with and without sonication so that the MZW the small capillary (1.55 mm), which was also used in this of primary nucleation can also be determined for continuously research, the influence of these two flow phenomena, conse- operated crystallizers. By non-invasive particle observation, it quently acoustic flow, on the residence time behavior could be 312 J Flow Chem (2021) 11:303–319 Fig. 3 Residence time sum curve, in dependency of the dimensionless residence time, without periphery for the CFI in the USU for silent (left) and son- icated (right) conditions neglected. This means that the pure convection model can be determined from the difference of the saturation and outlet applied in the work of Valitov et al. to describe the residence temperature (ΔT) or from the inlet concentration and the out- − 1 time behavior of the sonicated straight tube of 1.55 mm inner let temperature. The mass flow rate was set to 17 g min . diameter. For the bigger capillary and longer sonication time Figure 4 shows the individually obtained supersaturations. (3.0 s) the influence of the acoustic streaming could be shown, A distinction is made between the sonicated and silent condi- so the dispersion flow model fitted better than the pure con- tions, and some of the individual experiments could be aver- vection model and indicated the presence of axial back mixing aged into one sample because of their similarity. The results due to the sonication, so the residence time behavior was are divided into three categories. Either no crystals were vis- influenced in this case [86]. ible the whole time, crystals were occasionally detected in the The results of residence time behavior in the CFI confirms transition range, or nucleation crystals were continuously vis- the investigations of Valito et al. [86] and would suggest that ible. Accordingly, crystal nuclei were repeatedly seen in the the acoustic flow in these capillaries (d 1.6 mm) can be transition area during the test, but there were also longer pe- neglected. Nevertheless, the small deviations of the actual riods (minutes) in which not a single crystal was seen under residence time and the Bodenstein number between the silent the microscope. If none of the trials would be assigned to the and sonicated conditions could be a hint that for small capil- transition range, there would be a supersaturation threshold laries and longer sonication times a small effect of the acoustic above which crystals could always be generated. If there is a streaming is present. transition range, then there are two boundaries: the first be- tween where no crystals are formed and stochastically crystals are formed and the second between stochastically crystals are Nucleation experiments formed and crystals are visible the whole time. Assuming that the nucleation behavior of a substance sys- The nucleation experiments were done under silent and soni- tem would be the same in a batch apparatus as in a cated conditions. To examine whether the nucleation in CFI under ultrasound exposure has a sharp limit or rather a transi- tion area, most of the supersaturations were in the area of 1.10 to 1.46. The supersaturation could not be completely reduced in any of the tests. Accordingly, the indicated supersaturation here always indicated the calculated supersaturation, which is Table 1 Key figures for the residence time of the CFI silent −1 m gmin T [°C] ts½ [s] Bo [-] 16.5 ± 0.8 23.3 ± 0.5 43.8 45.3 ± 0.7 161.5 ± 4.8 sonicated −1 m gmin T [°C] ts½ [s] Bo [-] Fig. 4 Nucleation in the USU under sonicated and silent conditions for different supersaturations; samples A – D are averaged values from sev- 16.5 ± 0.8 23.3 ± 0.5 43.8 44.8 ± 0.1 167.1 ± 6.9 eral experiments at almost the same supersaturation J Flow Chem (2021) 11:303–319 313 continuously flowed crystallizer, the following experimental cavitation bubbles are size distributed, which in turn leads effort between the different modes would result. Each exper- to size distribution of crystal nuclei. imental point is run over at least 14 to 20 minutes, with a When interpreting the results, it should be considered that residence time of 45 s. This results in about 18 to 26 experi- the influence of a different residence time and, consequently, a mental runs. If this were done within a batch experiment, the longer possible induction time has not been investigated, be- cooling rates would have to be run in again 18 to 26 times, and cause of the given boundary conditions (mass flow rate and after 45 s each time it would have to be determined whether manageable coiling length of the tube). If the residence time in nucleation occurred or not. This shows how much experimen- the ultrasonic bath could have been extended by a longer tube, tal effort can be saved by this newly introduced methodology. Fig. 4 would probably have changed that crystals also have First, for the nucleation experiments without ultrasound, been seen at lower supersaturations or the transition range i.e. under quiet conditions, it can be seen that no crystals were would have been smaller, since the possible induction time visible, neither for the smaller supersaturations in the range of could be longer. 1.10 to 1.33 (ΔT = 10 K to 34.1 K) and a saturated starting Since eight experiments were performed with approximate- solution at T = 50 °C (filled grey circles) nor for the higher ly the same supersaturation of 1.25, these experiments were supersaturations 1.42 and 1.46 (ΔT=34 K to41.3K)witha combined as sample A, and shown as a cross with a green saturated starting solution at T = 60 °C (empty circles). background (Fig. 4). This means that the S and S of these T G For a continuous stirred tank, which is especially construct- tests are averaged as well as their mass flow rate, ambient ed for nucleation of the system used, it is known from the temperature T and the determined and calculated pres- ambient, literature [10] that nucleation takes place from subcooling of sure loss. These values are listed in Table 2. This averaged about 2 K (46 °C to 44 °C, or supersaturation of 1.02) with an sample A consists of one experiment, in which no crystals induction time of 1.68 min. It should also be considered that were visible, three experiments, in which crystals were par- for continuously operated crystallizers, not only the MZW can tially visible, and four experiments with sustained nucleation. be considered, but also the residence time. Thus, the induction This leads to the conclusion that for the selected material sys- time must be considered and the MZW should therefore be tem and the apparatus arrangement the transition to a most called a primary nucleation threshold (PNT) [10]. Another probable formation of crystal nuclei is present from supersat- study investigated the MZW for different saturation and nu- uration of 1.249 ± 0.005. cleation temperatures for L-alanine/water in a batch stirred For other supersaturations in the nucleation range, experi- vessel [87]. Extrapolation of these data results in a required ments could be combined into one to compare these points supersaturation of about 1.4 for an induction time of around with each other in a more targeted way. For this purpose, at 45 s. least three experiments were averaged, which are also listed in The results of the silent nucleation tests suggest that good Table 2 with their associated process variables. In the diagram, mixing, produced by the flow in the CFI design, is not suffi- these samples B – D are also shown as blue symbols. In cient to produce nucleation at lower supersaturations because summary, it is shown as described before that lower supersat- even at a supersaturation of 1.46 there were no nuclei visible. urations are required for nucleation by sonication than with- Presumably, the PNT for silent conditions is even wider. out. Further, it was found that there is a transition area, in However, this was not considered further, since Eder et al. which nucleation takes place in the selected apparatus and [11] already found that for the generation of seed crystals in chosen model system. For the selected conditions it is found a tube without ultrasound, the system usually clogs. that the PNT starts at a supersaturation of 1.2 and above a Every single experiment under sonication is marked as x supersaturation of 1.26 crystallization always occurs. and sorted into the categories for the different In comparison to the data from the literature [42, 52, 58], (temperature) supersaturations in Fig. 4. Considering this these results could indicate that using lower frequencies in the arrangement, it was found that in a supersaturation range of rage of 20–40 kHz, a critical nucleation threshold in a super- 1.20–1.25 individual experiments are sorted into all three saturation range of 1.25 to 1.3 is present. To verify this state- categories. Accordingly, PNT was significantly reduced by ment, more investigations should be carried out with the same sonication of the tube. No special effect could be deter- equipment and the same energy input for many different sub- mined regarding the experimental days or the ambient tem- stance systems. perature. Therefore, the results show that under the select- The results also confirm the hypothesis that nucleation is − 1 ed process conditions of flow rate (17 g min )and used caused by the cavitation bubble rather than by the acoustic experimental setup there is no precisely defined supersat- flow, because the frequency range used is known for pro- uration threshold above which nucleation is generated by nounced cavitation bubble formation [16]. Additionally, ultrasound, but rather a transition region. This could be Valitov et al. [86] showed that acoustic flow is rather negligi- because both nucleation and the formation of cavitation ble for this capillary size, and even with an already very well- mixed system, generated by the tube design, the PNT could bubbles are stochastic processes. It is also known that 314 J Flow Chem (2021) 11:303–319 Table 2 Averaged nucleation experiments in the USU under sonicated conditions for different supersaturations −1 number of experimental S [-] m gmin T status Δp Δp [mbar] Δp [mbar] Δp ambient exp calc calc ̅calc runs [°C] [mbar] with f with f [mbar] a b A 8 1.249±0.005 17.2±0.3 23.5±1.0 transition 195.5±6.8 181.1 214.7 197.9 range B 3 1.270±0.005 17.3±0.4 23.5±0.3 nucleation 199.5±6.9 184.9 218.8 201.9 C 4 1.284±0.005 17.3±0.3 22.9±0.9 nucleation 206.5±6.6 188.9 222.5 205.7 D 4 1.298±0.003 17.3±0.3 23.2±0.9 nucleation 213.1±8.7 190.6 224.5 207.6 not be reduced. However, no cavitation bubble formation Pressure drop and clogging could be observed in a smaller capillary (d = 0.8 mm) without air slugs in the investigations of Sarac et al. [34]. In order to The experimentally determined pressure loss Δp is com- exp ensure that cavitation bubble formation is responsible for nu- pared to the calculated one. Figure 6 shows the measured cleation, it should be investigated if cavitation bubbles are pressure loss (filled rhombus), the calculated pressure loss formed in the capillary size of 1.6 mm. coefficient f (squares) and f (triangles) and the averaged a b pressure loss from both (plus sign) over the Reynolds number. It can be seen that the actual pressure loss with loss coefficient Crystal size distribution f calculated is too low and f is overestimated. Averaging the a b two calculated pressure losses gives a better prediction of the To compare the samples A – D, the relating CSD was also pressure loss for the range investigated here. This calculation investigated. In addition, another single experiment with su- can help to follow an automated operation by starting a flush- persaturation of 1.38 was added for evaluation. Photos were ing cycle in case of deviating pressure loss. taken at random times during the experiments, at least with a For most of the nucleation experiments, the time of clog- time interval of three seconds to ensure that no crystals are ging was also recorded, shown in Fig. 7. The shortest test photographed twice. Crystals with a diameter of less than duration was thus about 14 min and a maximum of 35 min. 10 µm were not included in the evaluation, as the image anal- Dependence of the supersaturation on the clogging time t clog ysis could not determine whether these were actually formed could not be detected, because both extrema could be found at crystal nuclei or only image artifacts. These random photos the same supersaturation. The mean time until clogging was were evaluated using the image evaluation tool described 22 min. Although a clogging is usually visible due to the above [77]. The results are shown in Fig. 5. The determined transparent tube, no location could be identified where clog- supersaturation is plotted as dependence of the diameter. This ging has repeatedly occurred. Additionally, it was not possible is the circular equivalent diameter. The number of optically to identify excessive solid loads, which could be the cause of evaluated particles is listed above each CSD. The crystal size clogging. It is assumed that the small particles (< 90 µm), distributions are drawn in classes (black points, with a class which tend to agglomerate more frequently [88], have caused width of 10 µm) and as boxplots. the clogging, although the homogeneous flow conditions are Based on the CSD results, no clear influence between su- fulfilled. Besides, the already formed crystal nuclei can lead to persaturation and the influence on crystal size can be identi- a strengthening of secondary nucleation, because not enough fied. For the lower supersaturations (< 1.3) a particularly larg- surface for the crystals to grow is available and, instead, sec- er amount of fine grain (< 25 µm) was detected. This can also ondary nucleation occurs, which leads to clogging due to the be seen in the small values for x . The mean crystal size is relatively high supersaturation. 10,0 between 24 and 49 µm. At the higher supersaturation single A similar experimental setup to ours can be found in the experiment, the particles are bigger and the x is 62 µm. work of Furuta et al. [89] with a pH swing crystallization. 50,0 However, since a very different number of crystals were eval- Here, a coiled tube made of PFA (d =2 mm)and alength uated in each experiment, an interpretation of the dependence of 20–40 m was inserted into a temperature-controlled ultra- − 1 between supersaturation and crystal size is not statistically sonic bath (40 kHz). The flow rate was set to 50 mL min . validated. For this reason, the only conclusion to be drawn However, the influence of ultrasound on the formation of nu- from these results is that crystals with a size of 10–150 µm clei was not investigated by the authors. The ultrasonic irradi- were produced, but the majority of them are below 100 µm. ation was only used to avoid clogging [89]. Therefore, the created crystals are in the same range of crystal In other research groups, too, observations have been made size as they were created by Eder et al. [11] with a different on clogging. For example, Eder et al. [11] could not achieve model system, but in most similar conditions. stable seed crystal formation without ultrasound. Likely, J Flow Chem (2021) 11:303–319 315 Fig. 5 Crystal size distribution, evaluated with image analysis, in dependence of the supersaturation (for averaged nucleation experiments (A-D) and one addi- tional experiment with higher su- persaturation (1.38)) direct clogging would also have occurred in our case if even here that the rinse cycles were scheduled after 10 min [13], i.e. higher supersaturations had been applied to produce crystals in a similar time range as our shortest run till clogging was without ultrasound. In the investigations of Han et al. [52]a observed. If stable nucleation operation without clogging is feasible, stable operation was defined, if 3 to 4 residence times guaranteed, the USU presented here can be used as a seed no clogging occurred. According to this definition, all exper- crystal generation unit for continuous crystallization equip- iments presented here would have run stably. However, this ment as known from [71, 72, 90–93]. cannot be compared directly because the diameter chosen by Han et al. is 4 mm [52]. But, their average velocity is in a − 1 comparable range from 8 to 16 cm s [52]toours ofaround − 1 13 cm s . They also found that supersaturations that are Conclusions punctual high and not uniformly distributed lead to clogging. In the case of Ezeanowi et al. [60], no clogging was observed Cooling crystallization was investigated within a capillary − 1 at similar average velocities of 5–9cm s and an internal coiled flow inverter (CFI) with a tube diameter of 1.6 mm diameter of 250 mm. The observed duration per experiment embedded in an ultrasonic bath for nucleation in a continuous was also 3–4 residence times [60]. flow with different supersaturations. For this purpose, the out- In combination with the results determined here, it can be let of the test facility has been equipped with a non-invasive assumed that clogging can be prevented with even higher flow cell under a microscope with a conventional photo cam- suspension flow, despite the previously mentioned process era. The CFI design ensures a very narrow residence time condition of homogeneous suspension. As an alternative to behavior of the liquid phase. At a mass flow rate of 16.5 g − 1 avoid clogging problems, a feed mixture with a solution and min , a Bodenstein number of approx. 161–167 was obtain- seed suspension can be used, or rinse cycles can be imple- ed with a residence time in the ultrasonic bath of 45 s. Only mented as Besenhard et al. [13] did. It is interesting to note very slight differences between the residence time under Fig. 7 Time of clogging after starting the process solution for the Fig. 6 Measured and calculated pressure loss in dependence of the respective supersaturation Reynolds number 316 J Flow Chem (2021) 11:303–319 Competing financial interest The authors declare no competing finan- sonication and quiet conditions could be detected which could cial interest. indicate an influence of the acoustic streaming. Under silent conditions, no nucleation was observed in the Abbreviations CFI, coiled flow inverter; CSD, crystal size distribution; investigated range of supersaturation from 1.10 to 1.46 for the FPU, feed andpumpunit; FEP, fluorinated ethylene propylene; MZW, model system L-alanine/water. The first nuclei could be de- metastable zone width; PFA, perfluoroalkoxy alkane; PNT, primary nu- cleation threshold; PEAs, Process Equipment Assemblies; PAU, product tected under sonication and supersaturation of 1.2. The results analysis unit; US, ultrasound; USU, ultrasonic unit of the nucleation experiments show that under the selected conditions (ultrasonic bath, CFI designed tube, L-alanine/wa- Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adap- ter, and residence time 45 s) the primary nucleation threshold tation, distribution and reproduction in any medium or format, as long as (PNT) could be reduced by ultrasound as expected. you give appropriate credit to the original author(s) and the source, pro- Additionally, no rigid threshold for nucleation could be found, vide a link to the Creative Commons licence, and indicate if changes were but a nucleation range was identified in the investigated setup. made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a Nucleation was detected for supersaturations larger than 1.25. credit line to the material. If material is not included in the article's The transition range indicates that the two stochastic processes Creative Commons licence and your intended use is not permitted by of cavitation bubble formation and nucleation must be suffi- statutory regulation or exceeds the permitted use, you will need to obtain ciently pronounced to ensure nucleation. No influence of the permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. set supersaturation could be determined on the crystal size distribution or the time until clogging. The minimum time until clogging occurred was 14 minutes. The crystal size is in a range from 10 to 150 µm, where the most are from the size References in the range of 24–49 µm after a residence time of 45 s. Overall, the investigated setup is suitable for the rapid 1 Ma Y, Wu S, Macaringue EGJ et al (2020) Recent progress in detection of the influence of supersaturation on material continuous crystallization of pharmaceutical products: precise prep- systems by sonication. This setup consisting of coiled aration and control. Org Process Res Dev. https://doi.org/10.1021/ acs.oprd.9b00362 polymer capillary within an ultrasound bath can easily be 2 Yazdanpanah N, Nagy ZK (2020) The Handbook of Continuous set up with existing laboratory equipment. The developed Crystallization. Royal Society of Chemistry, Cambridge measuring cell from a glass jacket, optical microscope, and 3 Melches C, Plate H, Schürhoff J et al (2020) The steps from digital camera could even enable to identify influences on batchwise to continuous crystallization for a fine chemical: a case crystal form quickly and easily by ultrasound. In addition, study. Crystals 10:542. https://doi.org/10.3390/cryst10060542 the ultrasonic unit (USU) can be used as a seed crystal 4 Zhang D, Xu S, Du S et al (2017) Progress of pharmaceutical con- tinuous crystallization. Engineering 3:354–364. https://doi.org/10. production unit and in combination with rinse protocols 1016/J.ENG.2017.03.023 for a combination of continuous crystallization for the 5 Bieringer T, Buchholz S, Kockmann N (2013) Future production small-scale with other crystal growth units. concepts in the chemical industry: modular - small-scale - continu- ous. Chem Eng Technol 36:900–910. https://doi.org/10.1002/ceat. Acknowledgements The authors would like to thank Prof. Dr. Gerhard Schembecker and Dr. Kerstin Wohlgemuth (TU Dortmund University, 6 Schaber SD, Gerogiorgis DI, Ramachandran R et al (2011) BCI Laboratory of Plant and Process Design) for their inspiring discus- Economic analysis of integrated continuous and batch pharmaceu- sions on the investigations. We would also like to thank the reviewer for tical manufacturing: a case study. Ind Eng Chem Res 50:10083– the diligent review and helpful comments. Furthermore, we thank our 10092. https://doi.org/10.1021/ie2006752 technician Carsten Schrömges for the technical support. 7 Sun (2010) Particle size specifications for solid oral dosage forms: a regulatory perspective. https://www.americanpharma Author contributions The manuscript was written through contributions ceuticalreview.com/Featured-Articles/36779-Particle-Size- of all authors. All authors have given approval to the final version of the Specifications-for-Solid-Oral-Dosage-Forms-A-Regulatory- manuscript. Perspective/. Accessed 7 Oct 2020 8 Jong EJ de (1982) Entwicklung von Kristallisatoren. Chem Ing Tech 54:193–202. https://doi.org/10.1002/cite.330540303 Funding Open Access funding enabled and organized by Projekt DEAL. This research was funded by the German Federal Ministry of Economic 9 Hohmann L (2019) Process and equipment design for small-scale Affairs and Energy (BMWi) and the Project Management Jülich (PtJ) as continuous crystallization, 1. Auflage. Schriftenreihe part of the ENPRO2.0 initiative (Ref. no. 03ET1528A). Additionally, we Apparatedesign, Band 11. Dr. Hut, München acknowledge financial support by Deutsche Forschungsgemeinschaft and 10 Lührmann M-C, Timmermann J, Schembecker G et al (2018) Technische Universität Dortmund/TU Dortmund University within the Enhanced product quality control through separation of crystalliza- funding programme Open Access Publishing. tion phenomena in a four-stage MSMPR cascade. Cryst Growth Des 18:7323–7334. https://doi.org/10.1021/acs.cgd.8b00941 11 Eder RJP, Schrank S, Besenhard MO et al (2012) Continuous Compliance with ethical standards Sonocrystallization of Acetylsalicylic Acid (ASA): control of crystal size. Cryst Growth Des 12:4733–4738. https://doi.org/10.1021/ Conflicts of interest/Competing interests No conflict of interest cg201567y J Flow Chem (2021) 11:303–319 317 12 Jiang M, Papageorgiou CD, Waetzig J et al (2015) Indirect 31 Lee J (2016) Handbook of ultrasonics and sonochemistry: ultrasonication in continuous slug-flow crystallization. Cryst Importance of Sonication and Solution Conditions on the Acoustic Growth Des 15:2486–2492. https://doi.org/10.1021/acs.cgd. Cavitation Activity. Springer, Singapore 5b00263 32 Wohlgemuth K, Ruether F, Schembecker G (2010) 13 Besenhard MO, Neugebauer P, Ho C-D et al (2015) Crystal size Sonocrystallization and crystallization with gassing of adipic acid. control in a continuous tubular crystallizer. Cryst Growth Des 15: Chem Eng Sci 65:1016–1027. https://doi.org/10.1016/j.ces.2009. 1683–1691. https://doi.org/10.1021/cg501637m 09.055 14 Wong SY, Cui Y, Myerson AS (2013) Contact secondary nucle- 33 Wohlgemuth K, Kordylla A, Ruether F et al (2009) Experimental ation as a means of creating seeds for continuous tubular crystal- study of the effect of bubbles on nucleation during batch cooling lizers. Cryst Growth Des 13:2514–2521. https://doi.org/10.1021/ crystallization. Chem Eng Sci 64:4155–4163. https://doi.org/10. cg4002303 1016/j.ces.2009.06.041 15 Ruecroft G, Hipkiss D, Ly T et al (2005) Sonocrystallization: the use 34 Sarac BE, Stephens DS, Eisener J et al (2020) Cavitation bubble of ultrasound for improved industrial crystallization. Org Process dynamics and sonochemiluminescence activity inside sonicated Res Dev 9:923–932. https://doi.org/10.1021/op050109x submerged flow tubes. Chem Eng Process Process Intensif 150: 16 Nalesso S, Bussemaker MJ, Sear RP et al (2019) A review on pos- 107872. https://doi.org/10.1016/j.cep.2020.107872 sible mechanisms of sonocrystallisation in solution. Ultrason 35 Hem SL, Skauen DM, Beal HM (1967) Mechanism of crystalliza- Sonochem 57:125–138. https://doi.org/10.1016/j.ultsonch.2019.04. tion of hydrocortisone by ultrasonic irradiation. J Pharm Sci 56:229– 233. https://doi.org/10.1002/jps.2600560216 17 Thompson LH, Doraiswamy LK (1999) Sonochemistry: science 36 Lee J, Ashokkumar M, Kentish SE (2014) Influence of mixing and and engineering. Ind Eng Chem Res 38:1215–1249. https://doi. ultrasound frequency on antisolvent crystallisation of sodium chlo- org/10.1021/ie9804172 ride. Ultrason Sonochem 21:60–68. https://doi.org/10.1016/j. 18 Edmonds P, Dunn D (1981) Introduction: physical description of ultsonch.2013.07.005 ultrasonic fields. In: Edmonds PD (ed) Methods of experimental 37 Zeiger BW, Suslick KS (2011) Sonofragmentation of molecular physics // Ultrasonics. Acad. Press, Academic, New York, pp 1–28 crystals. J Am Chem Soc 133:14530–14533. https://doi.org/10. 19 Pohl B (2012) Charakterisierung der Mischung und Fällung bei 1021/ja205867f kontinuierlichen sonochemischen Reaktoren unter besonderer 38 Jordens J, Appermont T, Gielen B et al (2016) Sonofragmentation: Berücksichtigung der Reaktorform. Dissertation, Technische effect of ultrasound frequency and power on particle breakage. Cryst Universität Bergakademie Freiberg Growth Des 16:6167–6177. https://doi.org/10.1021/acs.cgd. 20 Apfel RE. Acoustic cavitation. In Peter D. Edmonds: Methods in 6b00088 experimental physics. Academic Press 19:355–411. https://doi.org/ 39 Jiang M, Gu C, Braatz RD (2017) Analysis of focused indirect 10.1016/S0076-695X(08)60338-5 ultrasound via high-speed spatially localized pressure sensing and 21 Kim H, Suslick K (2018) The effects of ultrasound on crystals: its consequences on nucleation. Chem Eng Process Process Intensif sonocrystallization and sonofragmentation. Crystals 8:280. https:// 117:186–194. https://doi.org/10.1016/j.cep.2017.04.008 doi.org/10.3390/cryst8070280 40 Nalajala VS, Moholkar VS (2011) Investigations in the physical 22 Brennen CE (2014) Cavitation and bubble dynamics. Cambridge mechanism of sonocrystallization. Ultrason Sonochem 18:345– University Press, Cambridge 355. https://doi.org/10.1016/j.ultsonch.2010.06.016 23 Leong T, Ashokkumar M, Kentish S (2011) The fundamentals of 41 Jordens J, Gielen B, Braeken L et al (2014) Determination of the power ultrasound - a review. Acoust Aust 39:54–63 effect of the ultrasonic frequency on the cooling crystallization of 24 Harzali H, Baillon F, Louisnard O et al (2011) Experimental study paracetamol. Chem Eng Process Process Intensif 84:38–44. https:// of sono-crystallisation of ZnSO4 7H2O, and interpretation by the doi.org/10.1016/j.cep.2014.01.006 segregation theory. Ultrason Sonochem 18:1097–1106. https://doi. 42 Jordens J, Canini E, Gielen B et al (2017) Ultrasound assisted par- org/10.1016/j.ultsonch.2011.03.007 ticle size control by continuous seed generation and batch growth. 25 Dodds J, Espitalier F, Louisnard O et al (2007) The effect of ultra- Crystals 7:195. https://doi.org/10.3390/cryst7070195 sound on crystallisation-precipitation processes: some examples and 43 Narducci O, Jones AG, Kougoulos E (2011) Continuous crystalli- a new segregation model. Part Part Syst Charact 24:18–28. https:// zation of adipic acid with ultrasound. Chem Eng Sci 66:1069–1076. doi.org/10.1002/ppsc.200601046 https://doi.org/10.1016/j.ces.2010.12.008 26 Grossier R, Louisnard O, Vargas Y (2007) Mixture segregation by 44 Li H, Wang J, Bao Y et al (2003) Rapid sonocrystallization in the an inertial cavitation bubble. Ultrason Sonochem 14:431–437. salting-out process. J Cryst Growth 247:192–198. https://doi.org/10. https://doi.org/10.1016/j.ultsonch.2006.10.010 1016/S0022-0248(02)01941-3 27 Kurotani M, Miyasaka E, Ebihara S et al (2009) Effect of ultrasonic 45 Guo Z, Zhang M, Li H et al (2005) Effect of ultrasound on anti- irradiation on the behavior of primary nucleation of amino acids in solvent crystallization process. J Cryst Growth 273:555–563. supersaturated solutions. J Cryst Growth 311:2714–2721. https:// https://doi.org/10.1016/j.jcrysgro.2004.09.049 doi.org/10.1016/j.jcrysgro.2009.03.009 46 Gracin S, Uusi-Penttilä M, Rasmuson ÅC (2005) Influence of ultra- 28 Miyasaka E, Kato Y, Hagisawa M et al (2006) Effect of ultrasonic sound on the nucleation of polymorphs of p -Aminobenzoic Acid. irradiation on the number of acetylsalicylic acid crystals produced Cryst Growth Des 5:1787–1794. https://doi.org/10.1021/cg050056a under the supersaturated condition and the ability of controlling the 47 Kordylla A, Koch S, Tumakaka F et al (2008) Towards an optimized final crystal size via primary nucleation. J Cryst Growth 289:324– crystallization with ultrasound: Effect of solvent properties and ul- 330. https://doi.org/10.1016/j.jcrysgro.2005.11.084 trasonic process parameters. J Cryst Growth 310:4177–4184. 29 Miyasaka E, Ebihara S, Hirasawa I (2006) Investigation of primary https://doi.org/10.1016/j.jcrysgro.2008.06.057 nucleation phenomena of acetylsalicylic acid crystals induced by ultrasonic irradiation—ultrasonic energy needed to activate primary 48 Louhi-Kultanen M, Karjalainen M, Rantanen J et al (2006) nucleation. J Cryst Growth 295:97–10 Crystallization of glycine with ultrasound. Int J Pharm 320:23–29. 1. https://doi.org/10.1016/j. https://doi.org/10.1016/j.ijpharm.2006.03.054 jcrysgro.2006.07.020 30 Hem SL (1967) The effect of ultrasonic vibrations on crystallization 49 Kurotani M, Hirasawa I (2008) Polymorph control of sulfamerazine processes. Ultrasonics 5:202–207. https://doi.org/10.1016/0041- by ultrasonic irradiation. J Cryst Growth 310:4576–4580. https:// 624X(67)90061-3 doi.org/10.1016/j.jcrysgro.2008.08.002 318 J Flow Chem (2021) 11:303–319 50 Hatakka H, Alatalo H, Louhi-Kultanen M et al (2010) Closed-loop flow inverter. Ind Eng Chem Res 59:3934–3942. https://doi.org/ control of reactive crystallization PART II: polymorphism control of 10.1021/acs.iecr.9b04160 L-glutamic acid by sonocrystallization and seeding. Chem Eng 70 Hohmann L, Gorny R, Klaas O et al (2016) Design of a continuous Technol 33:751–756. https://doi.org/10.1002/ceat.200900577 tubular cooling crystallizer for process development on lab-scale. 51 Nguyen T, Khan A, Bruce L et al (2017) The effect of ultrasound on Chem Eng Technol 39:1268–1280. https://doi.org/10.1002/ceat. the crystallisation of paracetamol in the presence of structurally similar impurities. Crystals 7:294. https://doi.org/10.3390/cryst7100294 71 Hohmann L, Greinert T, Mierka O et al (2018) Analysis of crystal 52 Han B, Ezeanowi NC, Koiranen TO et al (2018) Insights into design size dispersion effects in a continuous coiled tubular crystallizer: criteria for a continuous, sonicated modular tubular cooling crystal- experiments and modeling. Cryst Growth Des 18:1459–1473. lizer. Cryst Growth Des 18:7286–7295. https://doi.org/10.1021/acs. https://doi.org/10.1021/acs.cgd.7b01383 cgd.8b00700 72 Schmalenberg M, Hohmann L, Kockmann N (2018) Miniaturized 53 Amara N, Ratsimba B, Wilhelm A-M et al (2001) Crystallization of tubular cooling crystallizer with solid-liquid flow for process devel- potash alum: effect of power ultrasound. Ultrason Sonochem 8:265– opment. In: Proceedings of the ASME 16th International 270. https://doi.org/10.1016/S1350-4177(01)00087-6 Conference on Nanochannels, Microchannels, and Minichannels 54 Ruecroft G, Bruns J (2010) An apparatus and process for producing 2018. American Society of Mechanical Engineers Digital crystals (WO 2010/079350 A2). Accessed 26 Aug 2020 Collection; The American Society of Mechanical Engineers, New 55 Kudo S, Takiyama H (2012) Production of fine organic crystalline York. https://doi.org/10.1115/ICNMM2018-7660 particles by using milli segmented flow crystallizer. J Chem Eng Jpn 73 Kurt SK, Gelhausen MG, Kockmann N (2015) Axial dispersion and 45:305–309. https://doi.org/10.1252/jcej.11we168 heat transfer in a milli/microstructured coiled flow inverter for nar- 56 Hussain MN, Jordens J, John JJ et al (2019) Enhancing pharmaceu- row residence time distribution at laminar flow. Chem Eng Technol tical crystallization in a flow crystallizer with ultrasound: Anti- 38:1122–1130. https://doi.org/10.1002/ceat.201400515 solvent crystallization. Ultrason Sonochem 59:104743. https://doi. 74 Hohmann L, Schmalenberg M, Prasanna M et al (2019) Suspension org/10.1016/j.ultsonch.2019.104743 flow behavior and particle residence time distribution in helical tube 57 Savvopoulos SV, Hussain MN, Jordens J et al (2019) A mathemat- devices. Chem Eng J 360:1371–1389. https://doi.org/10.1016/j.cej. ical model of the ultrasound-assisted continuous tubular crystalliza- 2018.10.166 tion of aspirin. Cryst Growth Des 19:5111–5122. https://doi.org/10. 75 Klose A, Merkelbach S, Menschner A et al (2019) Orchestration 1021/acs.cgd.9b00466 requirements for modular process plants in chemical and pharma- 58 Rossi D, Jamshidi R, Saffari N et al (2015) Continuous-flow ceutical industries. Chem Eng Technol 42:2282–2291. https://doi. sonocrystallization in droplet-based microfluidics. Cryst Growth org/10.1002/ceat.201900298 Des 15:5519–5529. https://doi.org/10.1021/acs.cgd.5b01153 76 Elma Schmidbauer GmbH (2020) Elmasonic S 30 H: Ultraschall- 59 Koiranen T, Ekberg B, Häkkinen A et al (2018) An ultrasound Reinigungsgerät. https://www.elma-ultrasonic.com/fileadmin/ crystallization device and an ultrasound crystallization system downloads/Produktprofile/Produktprofile_DE/Ultraschallgeraete/ (WO/2018/096205). Accessed 26 Aug 2020 Elmasonic_S/PP_Elmasonic_S30H_DE.pdf. Accessed 14 Sep 2020 60 Ezeanowi N, Pajari H, Laitinen A et al (2020) Monitoring the dynam- 77 Schmalenberg M, Sallamon F, Haas C et al (2020) Temperature- ics of a continuous sonicated tubular cooling crystallizer. Cryst controlled minichannel flow-cell for non-invasive particle measure- Growth Des 20:1458–1466. https://doi.org/10.1021/acs.cgd.9b01103 ments in solid-liquid flow. In: Proceedings of the ASME 18th 61 Zierep J, Bühler K (2018) Grundzüge der Strömungslehre: International Conference on Nanochannels, Microchannels, and Grundlagen, Statik und Dynamik der Fluide, 11. Aufl. 2018. Minichannels 2020. American Society of Mechanical Engineers Springer Fachmedien Wiesbaden; Imprint: Springer Vieweg, Digital Collection; The American Society of Mechanical. https:// Wiesbaden doi.org/10.1115/ICNMM2020-1062 62 Saxena AK, Nigam KDP (1984) Coiled configuration for flow in- 78 Borchert C, Sundmacher K (2011) Crystal aggregation in a flow version and its effect on residence time distribution. AIChE J 30: tube: image-based observation. Chem Eng Technol 34:545–556. 363–368. https://doi.org/10.1002/aic.690300303 https://doi.org/10.1002/ceat.201000465 63 Klutz S, Kurt SK, Lobedann M et al (2015) Narrow residence time 79 Huo Y, Liu T, Liu H et al (2016) In-situ crystal morphology iden- distribution in tubular reactor concept for Reynolds number range of tification using imaging analysis with application to the L-glutamic 10–100. Chem Eng Res Des 95:22–33. https://doi.org/10.1016/j. acid crystallization. Chem Eng Sci 148:126–139. https://doi.org/10. cherd.2015.01.003 1016/j.ces.2016.03.039 64 Mansour M, Liu Z, Janiga G et al (2017) Numerical study of liquid- 80 Kumar V, Mridha M, Gupta AK et al (2007) Coiled flow inverter as liquid mixing in helical pipes. Chem Eng Sci 172:250–261. https:// a heat exchanger. Chem Eng Sci 62:2386–2396. https://doi.org/10. doi.org/10.1016/j.ces.2017.06.015 1016/j.ces.2007.01.032 65 Sharma L, Nigam K, Roy S (2017) Single phase mixing in coiled 81 (2013) VDI-Wärmeatlas. Springer Berlin Heidelberg, Berlin tubes and coiled flow inverters in different flow regimes. Chem Eng 82 Emig G, Klemm E (2017) Chemische Reaktionstechnik. Springer Sci 160:227–235. https://doi.org/10.1016/j.ces.2016.11.034 Berlin Heidelberg, Berlin 66 Vashisth S, Nigam KDP (2008) Liquid-phase residence time distri- 83 Gelhausen MG, Kurt SK, Kockmann N (2014) Mixing and heat bution for two-phase flow in coiled flow inverter. Ind Eng Chem transfer in helical capillary flow reactors with alternating bends. Res 47:3630–3638. https://doi.org/10.1021/ie070447h In: ASME 2014 12th International Conference on Nanochannels, 67 Rossi D, Gargiulo L, Valitov G et al (2017) Experimental charac- Microchannels, and Minichannels collocated with the ASME 2014 terization of axial dispersion in coiled flow inverters. Chem Eng Res 4th Joint US-European Fluids Engineering Division Summer Des 120:159–170. https://doi.org/10.1016/j.cherd.2017.02.011 Meeting. ASME 68 Kurt SK, Akhtar M, Nigam KDP et al (2017) Continuous reactive precipitation in a coiled flow inverter: inert particle tracking, modu- 84 Wohlgemuth K, Schembecker G (2013) Modeling induced nucle- lar design, and production of uniform CaCO 3 particles. Ind Eng ation processes during batch cooling crystallization: A sequential Chem Res 56:11320–11335. https://doi.org/10.1021/acs.iecr. parameter determination procedure. Comput Chem Eng 52:216– 7b02240 229. https://doi.org/10.1016/j.compchemeng.2012.12.001 69 Benitez-Chapa AG, Nigam KDP, Alvarez AJ (2020) Process inten- 85 Lorimer JP, Mason TJ, Fiddy K (1991) Enhancement of chemical si reactivity by power ultrasound: an alternative interpretation of the fication of continuous antisolvent crystallization using a coiled J Flow Chem (2021) 11:303–319 319 Lena K. Weick studied hot spot. Ultrasonics 29:338–343. https://doi.org/10.1016/0041- Chemical Engineering at TU 624X(91)90032-4 DortmundUniversity since 2013. 86 Valitov G, Jamshidi R, Rossi D et al (2020) Effect of acoustic stream- She graduated with the master ing on continuous flow sonocrystallization in millifluidic channels. thesis “Investigation ofdifferent Chem Eng J 379:122221. https://doi.org/10.1016/j.cej.2019.122221 operating modes of a coiled tube 87 Durga KKH, Selvarajan P, Shanthi D (2012) Nucleation kinetics, cooling crystall izer ” at XRD and SHG studies of L-Alanine single crystals grown at differ- theLaboratory of Equipment ent supersaturation levels. Int J Curr Res Rev 2012:68–77 Design in 2020. 88 Hofmann G (ed) (2004) Kristallisation in der industriellen Praxis. Wiley-VCH, Weinheim 89 Furuta M, Mukai K, Cork D et al (2016) Continuous crystallization using a sonicated tubular system for controlling particle size in an API manufacturing process. Chem Eng Process Process Intensif 102:210–218. https://doi.org/10.1016/j.cep.2016.02.002 90 Eder RJP, Radl S, Schmitt E et al (2010) Continuously seeded, continuously operated tubular crystallizer for the production of ac- tive pharmaceutical ingredients. Cryst Growth Des 10:2247–2257. https://doi.org/10.1021/cg9015788 Norbert K ockman n 91 Eder RJP, Schmitt EK, Grill J et al (2011) Seed loading effects on studiedmechanical engineering at the mean crystal size of acetylsalicylic acid in a continuous-flow the Technical University of crystallization device. Cryst Res Technol 46:227–237. https://doi. Munich and completed hisDr.- org/10.1002/crat.201000634 Ing. in process engineering at the 92 Wiedmeyer V, Anker F, Bartsch C et al (2017) Continuous crystal- University in Bremen. After 5 lization in a helically coiled flow tube: analysis of flow field, resi- years in chemicalindustry as pro- dence time behavior, and crystal growth. Ind Eng Chem Res 56: ject manager he joined Freiburg 3699–3712. https://doi.org/10.1021/acs.iecr.6b04279 University, IMTEK in 2001 93 Wiedmeyer V, Voigt A, Sundmacher K (2017) Crystal population asgroup leader for micro process growth in a continuous helically coiled flow tube crystallizer. Chem engineering. In 2007, Dr. Eng Technol 40:1584–1590. https://doi.org/10.1002/ceat.201600530 Kockmann joined LonzaLtd., Visp, Switzerland, as senior scien- tist responsible for continuous flowprocesses and microreactor Publisher’snote Springer Nature remains neutral with regard to jurisdic- technology. In April 2011, Norbert Kockmann wasappointed as full pro- tional claims in published maps and institutional affiliations. fessor for equipment design at TU Dortmund University,Germany. His research interests are on small-scale device for continuouschemical pro- cesses, modular design, and process intensification. His workincludes Mira Schmalenberg studied fundamental investigations of small-scale multiphase flow, modellingand Chemical Engineering at TU simulation accompanied by modern sensing technology and machine- DortmundUniversity. In November learningmethods. 2017, she completed her master the- sis on the design ofa modular con- tinuous-flow tubular cooling crys- tallizer for process developmentand small-scale production at the Laboratory of Equipment Design. In February2018, she started as a research assistant in the ENPRO2.0-TeiA(‘‘Trennverfahren mit effizienten und intelligenten Apparat en ’’ ) project at theLaboratory of Equipment Design. Her focus is theinvestigation of small scale continuous cooling crys- tallizers.

Journal

Journal of Flow ChemistrySpringer Journals

Published: Sep 1, 2021

Keywords: Coiled flow inverter; Continuous nucleation; Cooling crystallization; Minichannel; Capillary flow; Non-invasive nucleation detection; Sonocrystallization

References