Research Article: Challenges of Mechanochemistry: Is In Situ Real‐Time Quantitative Phase Analysis Always Reliable? A Case Study of Organic Salt Formation

Date Published: May 08, 2017

Publisher: John Wiley and Sons Inc.

Author(s): Adam A. L. Michalchuk, Ivan A. Tumanov, Sumit Konar, Simon A. J. Kimber, Colin R. Pulham, Elena V. Boldyreva.


Mechanochemical methods offer unprecedented academic and industrial opportunities for solvent‐free synthesis of novel materials. The need to study mechanochemical mechanisms is growing, and has led to the development of real‐time in situ X‐ray powder diffraction techniques (RI‐XRPD). However, despite the power of RI‐XRPD methods, there remain immense challenges. In the present contribution, many of these challenges are highlighted, and their effect on the interpretation of RI‐XRPD data considered. A novel data processing technique is introduced for RI‐XRPD, through which the solvent‐free mechanochemical synthesis of an organic salt is followed as a case study. These are compared to ex situ studies, where notable differences are observed. The process is monitored over a range of milling frequencies, and a nonlinear correlation between milling parameters and reaction rate is observed. Kinetic analysis of RI‐XRPD allows, for the first time, observation of a mechanistic shift over the course of mechanical treatment, resulting from time evolving conditions within the mechanoreactor.

Partial Text

One of the most challenging factors in RI‐XRPD is the associated artificial peak broadening and stochastic fluctuation of scattering intensities. In an ex situ experiment, the sample is immobile and its position with respect to the X‐ray beam is well defined and calibrated. By contrast, during an in situ experiment, the sample is in permanent motion within the milling jar, and thus the path of the immobile beam through the sample is time‐dependent. The stochastic motion of milling bodies within the milling jar further complicates in situ measurements. Hence, the sample that is analyzed by X‐ray diffraction varies not only because of a physical or chemical transformation, but also due to the stochastic motion of different particles flying in and out of the beam. While one can roughly estimate peak broadening to account for the thickness of the sample (with the maximum possible value determined by the jar diameter),17 its exact value is dynamic and dependent on the quantity and position of powder in the beam path at any moment in time. The ratio of phases present in this stochastically sampled material does not necessarily represent the composition of the sample as a whole. Such effects have been noted where the formation of different phases occurs at different jar sites.9 This proves very challenging for the processing of diffraction data. Of further complication is the considerable background associated with the milling jar, and the associated low signal‐to‐background intensity. Together, these issues render data processing a challenge, particularly by automated refinement strategies. Current methodologies for treating RI‐XRPD data17, 23 typically involve addition of a nonreactive calibrant into the sample, which is used to normalize and correct the data. In principle, this accounts for fluctuations in the quantity of diffracting sample, and either the ratio of sample to calibrant intensities, Ia/Ical, or automated Rietveld refinement (ARR) used to quantify the phase content and identify amorphous material. However, a number of issues surround this technique: (1) perfect mixing of the calibrant throughout the powder mixture must be assumed, and any deviation can be a source of erroneous amorphous content assignment, (2) the calibrant must be assumed not to act as a milling body, affect the rheology of the mixture, or to have any other chemical/physical effect on the milled powder, and (3) the use of Ia/Ical does not account for distribution of diffracted intensity across a broad signal. Even applying ARR in the absence of calibrant, while robust in many cases, encounters issues when faced with stochastic fluctuations in peak‐profile parameters, abnormal backgrounds, artificial peak splitting, or otherwise abnormally shaped peaks.17 Indeed, this led to issues in the processing of the present data, where phase fractions of the product phase were overestimated in the initial stages of the process and the profile dynamics were poorly represented in some cases. Instead, we suggest a hybrid technique (HT), combining careful Rietveld refinement of selected diffraction patterns and peak integration. We note that during milling, the powder is in continuous motion, and preferred orientation is negligible. This has been confirmed through Rietveld refinement. This method is summarized in Figure1Ia, and details are given in the Experimental Section. This type of methodology is expected to offer a general means by which difficult or fluctuating peak shapes may be treated, and can be extended to systems in which peak splitting is observed. Its main requirement is the existence of a high‐intensity, well‐resolved diffraction peak for each phase.

In Situ Milling: Real time in situ milling experiments were conducted at the European Synchrotron Radiation Facility (ESRF), beam line ID11, experiment CH4313. Ball milling was done in a modified MM400 Retsch mill. For each reaction, 300 mg of stoichiometric mixture of oxalic acid dihydrate and glycine was used. Perspex milling jars (14.5 mL) were used34 with a single stainless‐steel ball (7 mm diameter). Monochromatic X‐ray of wavelength 0.141696 Å was used, and powder patterns were collected every 0.4 s. Data were averaged by summing 10 detector frames, giving a total time resolution of 4 s. Integration of 2D data was performed using the PyFAI azimuthal integration methodology.

The authors declare no conflict of interest.




Leave a Reply

Your email address will not be published.