Oscillatory Baffled Crystallizer - CMAC

Stirred tank crystallizer– Strong mixing at impeller tips dissipated quickly– Poor heat and mass transfer due to ineffectivemixing overall– Complex scale upRistic, R.I., Chem. Eng. Res. Des., 2007. 85(7): p. 937-944

2. Experimental set up & calibration– A simple polarimeter was constructed for analysisof product crystals– Heating and cooling provided by external waterbath

Product analysis– Previous research showed that nucleation mechanismcould retrospectively be related to handedness of productcrystals 1– Polarimetry enables analysis of crystal enantiomorphism– This is the methodology used throughout the study– On average 3100 crystals were analysed and counted1. Denk, E.G. and G.D. Botsaris, J. Cryst. Growth, 1972. 13/14: p. 493-499

Procedure– Solution saturated at 31 °C (approximately 10 M NaClO 3solution)– Solution cooled to 30 °C– Supersaturation at T = 1 is approximately 1.01– On reaching 30 °C, solution seeded with single seedcrystal for 3 minutes– Mixing stopped and seed withdrawn– Crystals developed overnight and analyzed

3. Results and discussions3.1 Benchmark testsTest 1 – With NO mixing as well as NO stirrer and baffles inboth a STC and an OBC100 % similarity to the seed was obtainedTest 2 – With NO mixing, but with the presence of stirrer andbaffles in both a STC and an OBC100 % similarity to the seed was obtainedSummary 1 – no spontaneous nucleation was found tooccur

Test 3 – Percentage similarity to seed with both mixing and thepresence of stirrer and baffles in both a STC and anOBCP/V (Wm-3)sample 12 187 766STC OBC STC OBC STC OBC1 100 94 100 99 100 93.22 100 95 100 93.4 100 93.13 100 96.3 100 92.9 100 98.3Average 100 95.1 100 95.1 100 94.87Standard Error 0 0.67 0 1.96 0 1.72

Summary 2– Results found that the STC always gave crystals of thesame form as the seed crystal.– In the OBC, the product crystals were never 100 % similarto the seed, typically around 95 %.– Changing the oscillation frequency had little significanteffect on the similarity.– What has caused this production of the opposite-to-seedhanded crystals in the OBC?

The OBC step up– Baffles are tightly fitted to the column, the scraping actioncould be the main culprit– If this was true, the removal of it from the OBC while adding itto the STC would cast some insights into this hypothesis

The STC setup– Scraping was introduced by simply lowering the impeller

3.2 Testing hypothesis of scraping– When the gap was introduced in the OBC, similarities increasedtowards 100 %– When the impeller was scraping the surface of the STC, opposite-toseedhanded crystals were generated.P/V (Wm-3)sample 12 187 766STC OBC STC OBC STC OBC1 92.6 100 100 98.8 93.6 99.62 94.6 97.0 99.0 99.5 96.3 99.83 99.7 99.9 92.2 98.7 97.3 99.0Average 95.6 99.0 97.09 99.0 95.76 99.5Standard Error 2.1 1.0 2.4 0.2 1.1 0.2

For the Scraped OBCScrapedAmplitude (mm)Sample 7 15 301 97.84 98.11 93.22 100 95.68 90.53 99.92 98.43 93.1Average 99.25 97.41 92.27Standard error 0.71 0.87 0.88Increasing amplitude generated more opposite-handed crystals

For the un-scraped OBCUn-scrapedAmplitude (mm)Sample 7 15 301 100 99.95 98.842 100 99.93 99.493 100 99.42 98.71Average 100 99.77 99.01Standard error 0 0.17 0.24Once again, increasing amplitude generated more oppositehandedcrystals, but effect is less pronounced

3.4 Nucleation rate– The number of crystals obtained in each experiment isknown.– Seeding was done for a fixed period of time for all testsbefore agitation was stopped, enabling a small number oflarge crystals.– The nucleation rates could be evaluated from thisinformation

– The amplitude is related to the mixing via the Strouhalnumber (St)St=D/(4πx o )– The nucleation rate of the opposite-to-seed handed crystalsmay be related to the Strouhal number, provided the otherparameters remained unchanged and independent:J’ = K’(St) m– K’ is the nucleation rate constant and m is the order

– For both the scraped and un-scraped OBCs, apower law plot of St vs. J’ can be generated.Scraped un-scrapedm -3.3 -2.2K’ 46 11r 2 0.98 0.8St

Summary 3– The Strouhal number has significant influence on the rateof nucleation of the opposite handed enantiomer.– Un-scraped system has a lower dependence on theStrouhal number than the scraped system.– Scraping has much higher nucleation rate than whenscraping was not present (around 10 x increase onscraping).

4. Closing remarks– In the OBC, scraping mechanism seems to be the maindriver for the opposite-to-seed nuclei being formed– When scraping is removed, the similarity to seed increases,but we still see the opposite-to-seed nuclei. Could this bedue to internal surface renewal as a function of mixing?– Can we utilise this to either promote wanted nucleation oravoid unwanted nucleation?

Thank you for your attention.Are there any answers?