A series of hitherto unknown mixed-metal phosphates of the monophosphate tungsten bronze structure family [MPTB; (WO3)2m(PO2)4] have been obtained by solution combustion synthesis followed by annealing (ϑ = 850°C) at appropriate oxygen pressures. These new phosphates show substitution of W5+ by either M3+1/3W6+2/3 (M: V, Cr, Fe, Mo) or Ti4+1/2W6+1/2. Members of the MPTB structural series with m = 2 [e.g. CrIII4/3WVI8/3O12(PO2)4; TiIV6/3WVI6/3O12(PO2)4] and m = 4 [e.g. Cr4/3W20/3O24(PO2)4] have been obtained. In the course of our investigation the crystal structure of WOPO4 (MPTB with m = 2: W4O12(PO2)4) has been re-determined from X-ray single-crystal data, showing monoclinic instead of the orthorhombic symmetry reported in literature (P21/m, Z = 1, 80 parameters, 1832 independent reflections R1 = 0.027, wR2 = 0.063). The crystal structures of MoIII4/3WVI8/3O12(PO2)4 and CrIII4/3WVI8/3O12(PO2)4 (MPTBs with m = 2) were also refined from single-crystal data {(Mo/W (Cr/W): P21/m, Z = 1, 80(86) parameters, 1782(1769) independent reflections, R1 = 0.035(0.059), wR2 = 0.081(0.146)}. These refinements indicate statistical distribution of MIII and WVI over the metal sites. By selected area electron diffraction the unit cell dimensions of CrIII4/3WVI8/3O12(PO2)4 and CrIII4/3WVI20/3O24(PO2)4 derived from XRPD and SXRD are confirmed. HRTEM images of Cr4/3W20/3O24(PO2)4 are in agreement with its assumed close structural relation to W8O24(PO2)4 and show an highly ordered atomic arrangement.

Epilepsy is a complex brain disorder characterized by repetitive seizure events. Epilepsy patients often suffer from various and severe physical and psychological comorbidities (e.g., anxiety, migraine, and stroke). While general comorbidity prevalences and incidences can be estimated from epidemiological data, such an approach does not take into account that actual patient-specific risks can depend on various individual factors, including medication. This motivates to develop a machine learning approach for predicting risks of future comorbidities for individual epilepsy patients. In this work, we use inpatient and outpatient administrative health claims data of around 19,500 U.S. epilepsy patients. We suggest a dedicated multimodal neural network architecture (Deep personalized LOngitudinal convolutional RIsk model—DeepLORI) to predict the time-dependent risk of six common comorbidities of epilepsy patients. We demonstrate superior performance of DeepLORI in a comparison with several existing methods. Moreover, we show that DeepLORI-based predictions can be interpreted on the level of individual patients. Using a game theoretic approach, we identify relevant features in DeepLORI models and demonstrate that model predictions are explainable in light of existing knowledge about the disease. Finally, we validate the model on independent data from around 97,000 patients, showing good generalization and stable prediction performance over time.