Structured ternary fluids as nanocrystal incubators for enhanced crystallization control

In crystallization from solution, a ubiquitous process in both industry and the natural world, nucleation is usually the rate-determining step, followed by faster crystal growth. Consequently, crystals typically exist in the nm-size range for such limited times that their investigation and manipulation is hindered. Here, we show that, owing to a degree of restricted diffusion, crystallization in structured ternary fluids (STFs) can proceed via higher nucleation rate and slower crystal growth pathways. This enables STFs to act as nanocrystal incubators, with the nanocrystals existing for extended times. We demonstrate that this generates enhanced crystallization control, with the three ambient pressure polymorphs of glycine, the α-, γ- and β-forms, all crystallizing from the octanol/ethanol/water STF, despite the well-known difficulty in crystallizing the slow growing γ-form and the instability of the β-form. The ability of STFs to produce notoriously hard to crystallize polymorphs should make them a versatile tool, ideal for polymorph discovery. This may enable a step change in the current, scatter-gun approach to polymorph screening. Furthermore, we show that aliquots of the nanocrystal-containing fluids can successfully seed metastable solutions. Hence, STFs may ultimately help provide a generic methodology for producing crystals and seed suspensions of any desired polymorph to supersede current targeted crystallization and seeding strategies.


Solubility of -glycine in the STFs
STFs containing ethanol, water and octanol were prepared for the required compositions on a 10 g scale. Excess -glycine was then weighed and added to each sample vial, which were then sealed using PTFE tape. Each STF composition was made twice, with each duplicate being stored in an oven set at 36°C and 50°C, respectively, for 4 months. After this time, the samples underwent hot filtration, and the residual glycine (which was all still α-glycine) was weighed and its mass recorded. From this, the amount of α-glycine that dissolved, and therefore the solubility of α-glycine in each sample, could be calculated.

Preparation of glycine-containing STFs
STFs containing ethanol, water and octanol at constant ethanol mass ratio of 0.40 were prepared on a 250 ml scale for the lower supersaturation -glycine crystallization experiments and on a 25 ml scale for the higher supersaturation -glycine experiments. The water mass ratio was varied from 0.10 to 0.55, so as to encompass the oil-rich, bicontinuous and water-rich regions, see Fig. 1a.
Typically, the required mass of water and glycine were mixed. Octan-1-ol was then added, followed by ethanol, to achieve the desired composition. These systems were stirred with a magnetic stirrer bar at 323 -333 K for 1 hour, or until all glycine had dissolved. For crystallizations at supersaturations corresponding to c/c sat of 2.00 or higher, where c sat and c are the -glycine saturation and actual concentrations, respectively, samples were sonicated in an ultrasound water bath for 2 hours, or until all glycine had dissolved.

Crystallizations in the STFs
Once all the glycine had dissolved, the samples were maintained at an elevated temperature above the saturation temperature for a further hour before cooling to 298 K. Cooling was achieved by either directly placing the samples in a water bath at 298 K (fast-cooling method) or cooling at a controlled rate of 10 K hour¯¹ in an oven (slow-cooling method). and glycine crystals were obtained in the STFs using both the fast-and slow-cooling methods, whereas -glycine crystals were obtained in the STFs using only the fast-cooling method. Crystals were extracted from solution via Büchner filtration after 1 week for and glycine and after 10-20 minutes for -glycine. STFextracted crystals were washed with the corresponding non-glycine containing STF, and then with ethanol to remove residual octanol. Samples were then left to dry at 333 K.

Seeding experiments
Separate 250 ml STFs were prepared that contained suspended nanocrystals of -, and glycine. For the and glycine suspended nanocrystals, STFs containing 0.25 mass fraction of water and 0.40 mass fraction of water, respectively, were used with a relative glycine concentration, c/c sat , of 1.30. These STFs were cooled via the slow-cooling method and were then left at 25 C for 16 hours for the nanocrystals to form. At this time, the STFs contained 1-5 sedimented crystals of size 0.1 mm but the majority of the glycine was still in the STF, either in solution or in the form of suspended nanocrystals.
For the -glycine suspended nanocrystals, STFs containing 0.25 mass fraction of water at a higher relative glycine concentration, c/c sat , of 2.10 were used alongside the fast-cooling method, with the -nanocrystals forming within 1 hour at 25 C, at which point, again, there were 1-5 sedimented crystals present.
For the seeding experiments, a 5.0 ml aliquot of the nanocrystal-containing STF was added to 15.0 ml of a binary solution of 0.40 mass fraction water and 0.60 mass fraction ethanol, which contained glycine at a supersaturation corresponding to c/c sat of 1.30. and glycine crystals were extracted for ATR-FTIR analysis from the seeded experiments, and the STFs used for the seeding, after 1 week. -glycine crystals were extracted for ATR-FTIR analysis from the seeded experiments, and the STFs used for the seeding, after 10-20 minutes, with the shortened timeframe being necessary due to the rapid solution-mediated transformation of -glycine into -glycine in unstructured solution.

ROY polymorph screening test
A binary system comprised of 0.475 mass fraction toluene and 0.525 mass fraction isopropanol (IPA), and an STF comprised of 0.100 mass fraction toluene, 0.375 mass fraction water and 0.525 mass fraction IPA, were made on a 20 g scale. ROY was then added to each sample to achieve a relative ROY concentration, c/c sat , of 5.5 at 25°C, with c sat = 30 mg cm -3 and 3.5 mg cm -3 for the binary system and STF, respectively. The binary and STF samples were sealed with PTFE tape before being heated to 65 °C on a hot plate with stirring until all the ROY was dissolved. Samples were then placed in an oven at 70 °C to ensure full dissolution of ROY. After 1 hour, the samples were removed from the oven, placed in a water bath at 25°C and left to crystallize undisturbed. The samples were carefully monitored and in the first trial, crystals were extracted 5-10 minutes after the first appearance of a mass of suspended ROY crystals in the STF. This gave the 4 ROY polymorphs Y, YN, ON and R in the STF. In subsequent tests on the STF system, crystals were extracted at different times to achieve the pure polymorphs. In particular, for the pure Y polymorph, crystals were extracted at earlier times as the Y polymorph grew the fastest (it is both the thermodynamic and kinetic product in the toluene-IPA system). The YN and ON polymorphs appeared next as a mass of suspended needle crystals, and at later times the suspended needle mass sedimented and then disappeared as the R polymorph crystallized. In the binary system, only Y crystals were obtained irrespective of the extraction time, consistent with the Y-polymorph being both the kinetic and thermodynamic product.

ATR-FTIR
Dry crystals were gently ground in a pestle and mortar. ATR-FTIR data was then obtained using a PerkinElmer Spectrum FT-IR spectrometer. A sample was placed on the diamond surface of the main instrument platform before being compressed using a metal anvil. The sample was scanned over a range of 4000 cm -1 -380 cm -1 .

Optical microscopy
Optical microscopy data were acquired on an Olympus BX50 microscope equipped with a digital camera. Samples were transferred to a clean glass microscope slide before images were taken. To assess crystal growth rates, standard glass microscope slides were cut in half before spreading silicone grease around the outside to form a border. 2-3 drops of fluid were then transferred to the centre of one of the prepared slides before sealing with the other half of the slide. The sealed sample was then placed on a Linkam heating/cooling block, which was fitted with a central hole to allow light transmission through the sample. The sample was heated to 333 K to ensure all glycine remained dissolved, before cooling to 298 K, where the temperature was held. After the first crystal was identified, images were taken once every minute, with the length and width of the crystal(s) in each sequential image then measured. The increase in circumference per unit time was calculated, and from this, the crystal growth rate was determined.

NMR diffusiometry
NMR diffusiometry techniques allow the estimation of diffusion coefficients by labelling spins spatially, allowing the spins to move, and then reading the label. Diffusion coefficients obtained using these methods can be used to detect restricted diffusion, albeit with the consideration that the boundaries are not rigid, as in the case of classic restricted diffusion studies in solid porous materials. 1 To conduct the investigation, we fixed all the parameters in a 13 C{ 1 H}-Doneshot45 pulse sequence, varied the composition of the STF mixture, and calculated the apparent diffusion coefficient.
A 500 MHz Varian spectrometer equipped with a probe able to deliver a nominal maximum pulsed field gradient of 62 G cm -1 was used to acquire Diffusion-Ordered SpectroscopY (DOSY) experiments at 298 K. The calibrated maximum gradient strength was 60.95 G cm -1 . Standard 5 mm NMR tubes were used. A capillary containing D 2 O was used to provide the lock signal.
Oneshot45 spectra 2 were acquired for diffusion measurements. ¹³C{ 1 H} diffusion measurements were taken with a diffusion delay (Δ) of 800 ms and diffusion gradient length (δ) of 3 ms. The gradient strength was incremented in 15 steps from 5.58 G cm -1 to 46.5 G cm -1 in equal steps of gradient squared. For 1 H diffusion measurements, the Δ was set to 400 ms and the δ was 2 ms. The gradient strength was incremented in 15 steps from 1.86 G cm -1 to 27.9 G cm -1 in equal steps of gradient squared. All results were analysed using mono-exponential fittings. To ensure there was no convection, the 1 H convection compensated double-stimulated echo DBPPSTE_cc pulse sequence 3,4 was used, with Δ = 200 ms and δ = 2 ms. The diffusion-encoding gradient pulses were set to 27.9 G cm -1 . The repetition time was 6.0 s, of which 4.0 s comprised the acquisition time. This was repeated with a diffusion delay of 400 ms. Further setup details are summarized in Table S2. All DOSY spectra were processed using VNMRJ 4.2 software.

Powder X-ray diffraction
To prepare each sample, the glycine crystals were ground gently to make a fine powder to reduce preferred orientation effects, before being spread evenly on a silicon powder specimen holder. Data were collected using a Bruker D8 Advance diffractometer operated at 40 kV and 40 mA. This produced Cu Kα1 and Kα2 radiation at a wavelength of 1.5406 Å. DIFFRAC.SUITE software was used to set up and control each experimental run. The width of the incident X-ray beam was 2, 4, or 6 mm depending on the quantity of the sample. A step size of 0.02° was used to scan 2θ values between 10° and 70°, holding the beam at each angle for 1.00, 1.67, and 2.50 s for the 6, 4 and 2 mm beams, respectively.     respectively. Note that characteristic PXRD peaks for -glycine are at 19.5and 29.7 (asterisked), -glycine has a characteristic PXRD peak at 18 (asterisked) and -glycine has a characteristic PXRD peak at 25.3 (asterisked). [5][6][7] The characteristic FTIR peaks for , , and -glycine at 909, 914 and 928 cm -1 , respectively. 5,8,9 The -glycine PXRD spectrum contains trace amounts of the -glycine polymorph, which can be most easily discerned by the tiny -glycine peaks at 14.8 and 19.5.