Named Entity Recognition and Normalization Applied to Large-Scale Information Extraction from the Materials Science Literature.

The number of published materials science articles has increased manyfold over the past few decades. Now, a major bottleneck in the materials discovery pipeline arises in connecting new results with the previously established literature. A potential solution to this problem is to map the unstructured raw text of published articles onto structured database entries that allow for programmatic querying. To this end, we apply text mining with named entity recognition (NER) for large-scale information extraction from the published materials science literature. The NER model is trained to extract summary-level information from materials science documents, including inorganic material mentions, sample descriptors, phase labels, material properties and applications, as well as any synthesis and characterization methods used. Our classifier achieves an accuracy (f1) of 87%, and is applied to information extraction from 3.27 million materials science abstracts. We extract more than 80 million materials-science-related named entities, and the content of each abstract is represented as a database entry in a structured format. We demonstrate that simple database queries can be used to answer complex "meta-questions" of the published literature that would have previously required laborious, manual literature searches to answer. All of our data and functionality has been made freely available on our Github ( https://github.com/materialsintelligence/matscholar ) and website ( http://matscholar.com ), and we expect these results to accelerate the pace of future materials science discovery.

Kinetics of non-equilibrium lithium incorporation in LiFePO4.

Lithium-ion batteries are a key technology for multiple clean energy applications. Their energy and power density is largely determined by the cathode materials, which store Li by incorporation into their crystal structure. Most commercialized cathode materials, such as LiCoO(2) (ref. 1), LiMn(2)O(4) (ref. 2), Li(Ni,Co,Al)O(2) or Li(Ni,Co,Mn)O(2) (ref. 3), form solid solutions over a large concentration range, with occasional weak first-order transitions as a result of ordering of Li or electronic effects. An exception is LiFePO(4), which stores Li through a two-phase transformation between FePO(4) and LiFePO(4) (refs 5-8). Notwithstanding having to overcome extra kinetic barriers, such as nucleation of the second phase and growth through interface motion, the observed rate capability of LiFePO(4) has become remarkably high. In particular, once transport limitations at the electrode level are removed through carbon addition and particle size reduction, the innate rate capability of LiFePO(4) is revealed to be very high. We demonstrate that the reason LiFePO(4) functions as a cathode at reasonable rate is the availability of a single-phase transformation path at very low overpotential, allowing the system to bypass nucleation and growth of a second phase. The Li(x)FePO(4) system is an example where the kinetic transformation path between LiFePO(4) and FePO(4) is fundamentally different from the path deduced from its equilibrium phase diagram.

Efficient band gap prediction for solids.

An efficient method for the prediction of fundamental band gaps in solids using density functional theory (DFT) is proposed. Generalizing the Delta self-consistent-field (ΔSCF) method to infinite solids, the Δ-sol method is based on total-energy differences and derived from dielectric screening properties of electrons. Using local and semilocal exchange-correlation functionals (local density and generalized gradient approximations), we demonstrate a 70% reduction of mean absolute errors compared to Kohn-Sham gaps on over 100 compounds with experimental gaps of 0.5-4 eV, at computational costs similar to typical DFT calculations.

Fatal poisoning in Nordic drug addicts in 2002.

The present study from 2002 includes medicolegally examined fatal poisonings among drug addicts in the five Nordic countries: Denmark, Finland, Iceland, Norway and Sweden. A common definition "drug addict" is applied by the participating countries. The number of deaths, age, sex, place of death, main intoxicant and other drugs present in the blood are recorded in order to obtain national data, as well as comparable Nordic data and data comparable to earlier studies from 1997 and 1991. The Icelandic results are commented on separately due to the low number of cases. The most fatal overdoses are seen in Norway, in both the death rate (number per 100,000 inhabitants=8.44) and in absolute number (n=232). The comparable figures for the other four countries are Denmark 5.43 (n=175), Iceland 3.6 (n=6), Finland 2.93 (n=94) and Sweden 2.56 (n=136). In earlier studies from 1991 and 1997, the highest death rate is seen in Denmark, with Norway as number two. Denmark is the only country where the death rate decreases from 1997 to 2002. A relatively large increase in deaths in the younger age groups (<30 years) is noted from 1997 to 2002, except in Denmark, where only a small increase in overdose deaths in very young people (15-19 years) is observed. Females account for 12-20% of the overdoses (three out of six deaths in Iceland). Relatively fewer deaths are recorded in the capital areas in 2002 than in 1997 and 1991, suggesting more geographically widespread drug use in the Nordic countries. Heroin/morphine is the single most frequently encountered main intoxicant, varying from 10% of the cases in Finland to 72% of the cases in Norway. Finland differs from the other countries in that a high percentage of the fatal overdoses in Finland are not caused by an illicit drug; buprenorphine overdoses are seen, and relatively few deaths resulting from heroin are seen. Methadone is the main intoxicant in 41% of the Danish overdose cases, 15% of the Norwegian cases, 4% of the Swedish cases and none of the Finnish overdose cases, an observation probably linked to different national prescription rules for methadone. The analytical screening reveals extended polydrug use. Frequently seen substances, in addition to the main intoxicant are amphetamine, tetrahydrocannabinol (THC), benzodiazepines and ethanol.

Drugged driving in the Nordic countries--a comparative study between five countries.

The purpose of this study was to compare whether the high incidence of drugged driving in Norway was different to that in the other Nordic countries. All blood samples received by Nordic forensic institutes during one week in 1996, from drivers suspected by the police of driving under the influence (Denmark: n = 255, Finland: n = 270, Iceland: n = 40, Sweden: n = 86, Norway: n = 149), were analysed for alcohol and drugs (benzodiazepines, cannabinoids, amphetamines, cocaine, opiates and a number of antidepressant drugs) independent of the primary suspicion, and using the same analytical cut-off levels at the different institutes. The primary suspicion was directed towards drugs in more than 40% of the Norwegian cases, drugs were detected in more than 70% of these samples. In only 0-3% of the cases from Denmark, Finland and Iceland, were drugs suspected, while the corresponding frequency for Sweden was 17%. However, evidential breath analyses were used for about three-quarters of the Swedish drivers suspected to be influenced by alcohol. Blood alcohol concentrations (BAC's) below the legal limits were found in 32, 18 and 2% of the Norwegian, Icelandic and Finnish cases, respectively (BAC < 0.05%), in 10% of the Danish cases (BAC < 0.08%) and in 20% of the Swedish cases (BAC < 0.02%). Drugs were most frequently found in the Norwegian and Swedish cases with no alcohol (80-83%). Similar frequencies of drugs in samples with BAC's above the legal limits (19-22%), were obtained for all countries. Benzodiazepines, tetrahydrocannabinol and amphetamine represented the most commonly detected drugs. Our results show that differences between Norway and other Nordic countries with regard to drugs and driving, are connected to the selection criteria made by the police and with more focus on drugged driving in Norway.

Drug addict deaths in the Nordic countries: a study based on medicolegally examined cases in the five Nordic countries in 1991.

The study includes medicolegally examined deaths among drug addicts in 1991 in the five Nordic countries: Denmark, Finland, Iceland, Norway and Sweden. A common definition of 'drug addict' was applied by the participating countries. The greatest number of drug addict deaths per 10(5) inhabitants was observed in Denmark followed, in descending order by Norway, Sweden, Finland and finally Iceland with only four deaths. The main difference between the countries was found in the number of fatal poisonings. The distribution according to geographical regions showed that about half of all drug addict deaths occurred in the metropolitan areas. Of the capitals, the greatest number of fatal poisonings per 10(5) inhabitants was seen in Oslo, followed by Copenhagen with a similar number, Stockholm with only the half, and Helsinki with a quarter. Heroin/morphine dominated as cause of death in fatal poisonings in Norway and Sweden. In Denmark, heroin/morphine caused about half of the fatal poisonings only, and nearly one third of the fatal poisonings was caused by methadone. Except for two cases in Sweden, methadone deaths were not seen in the other Nordic countries. Amphetamine caused one tenth of the fatal poisonings in Sweden. In Finland only one tenth of the deaths were caused by heroin/morphine and more by codeine, ethylmorphine and different drugs and poisons not classified in Single Convention on Narcotic Drugs 1961 or the International Convention on Psychotropic Substances 1971. A widespread use of alcohol, cannabis and benzodiazepines, diazepam especially, was seen in all the countries.

Fatal poisonings in young drug addicts in the Nordic countries: a comparison between 1984-1985 and 1991.

Fatal poisonings among young drug addicts (15-34 years) in the five Nordic countries, Denmark, Finland, Iceland, Norway and Sweden in 1991 were investigated and compared to a similar investigation for 1984-1985 (Sweden for 1984 only). A common definition of 'drug addict' has been applied by the participating countries. In both investigations, the greatest number of drug addict deaths was seen in Denmark calculated per 10(5) inhabitants, followed in descending order by Norway, Sweden, Finland and Iceland. An increased number of deaths was observed from 1984-1985 to 1991 in all five countries. The increase in Denmark and Sweden was small while the number of deaths was more than doubled in Norway and Finland. The increased number of cases in Norway and Sweden in 1991 is mainly due to a greater number of deaths in the age group 25-34 years. In Finland, the increased number was seen mainly in the age group 15-24 years. In the two investigations heroin/morphine caused most of the fatal poisonings in Norway and Sweden. In Denmark, heroin/morphine caused about half of the fatal poisonings only, and strong analgesics other than heroin/morphine caused about one third of the deaths. In 1984-1985 it was methadone, propoxyphene and ketobemidone and in 1991 mostly methadone. The number of heroin/morphine related deaths in Finland increased from 1984-1985 to 1991, but other drugs and poisons caused a much higher proportion of the deaths. Pentobarbital caused the only fatal poisoning in Iceland in 1991.

Fatal poisoning in drug addicts in the Nordic countries.

The study includes medicolegally examined fatal poisonings among drug addicts in 1997 in the five Nordic countries: Denmark, Finland, Iceland, Norway and Sweden, and the results are compared to a similar investigation from 1991. A common definition of "drug addict" was applied by the participating countries. The highest death rate by poisoning in drug addicts was observed in Denmark, where it was 6.54 per 10(5)inhabitants, followed by Norway with 6.35, Sweden with 2.21, Finland with 1.63 and Iceland with 1.20 per 10(5)inhabitants. All countries showed a higher death rate in 1997 than in 1991. For all countries the distribution of deaths according to geographical regions showed a decreasing number of drug deaths in the metropolitan area and an increasing number in other cities. Heroin/morphine dominated as the cause of death and was responsible for about 90% of the cases in Norway. In Sweden and Denmark, however, heroin/morphine caused only about 70% of the fatal poisonings. About 30% of the fatal poisonings in Denmark and Sweden were caused by other group I drugs, in Denmark mainly methadone and in Sweden mainly propoxyphene. Apart from two cases in Sweden methadone deaths were not seen in the other Nordic countries. In Finland heroin/morphine deaths have increased from about 10% in 1991 to about 40% in 1997. Forty-four percent of the fatal poisonings in Finland were caused by other group I drugs, mainly codeine and propoxyphene. The two fatal poisonings in Iceland were caused by carbon monoxide. Only few deaths in this investigation were caused by amphetamine and cocaine. A widespread use of alcohol, cannabis and benzodiazepines, especially diazepam, was seen in all the countries.

Low intensity conduction states in FeS2: implications for absorption, open-circuit voltage and surface recombination.

Pyrite (FeS2), being a promising material for future solar technologies, has so far exhibited in experiments an open-circuit voltage (OCV) of around 0.2 V, which is much lower than the frequently quoted 'accepted' value for the fundamental bandgap of ∼0.95 eV. Absorption experiments show large subgap absorption, commonly attributed to defects or structural disorder. However, computations using density functional theory with a semi-local functional predict that the bottom of the conduction band consists of a very low intensity sulfur p-band that may be easily overlooked in experiments because of the high intensity onset that appears 0.5 eV higher in energy. The intensity of absorption into the sulfur p-band is found to be of the same magnitude as contributions from defects and disorder. Our findings suggest the need to re-examine the value of the fundamental bandgap of pyrite presently in use in the literature. If the contribution from the p-band has so far been overlooked, the substantially lowered bandgap would partly explain the discrepancy with the OCV. Furthermore, we show that more states appear on the surface within the low energy sulfur p-band, which suggests a mechanism of thermalization into those states that would further prevent extracting electrons at higher energy levels through the surface. Finally, we speculate on whether misidentified states at the conduction band onset may be present in other materials.

Fatal poisoning in drug addicts in the Nordic countries in 2007.

The frequency of medico-legally examined fatal poisonings in 2007 among drug addicts was investigated in five Nordic countries; Denmark, Finland, Iceland, Norway, and Sweden. The number of deaths, age, sex, place of death, main intoxicant, and other drugs present in blood samples were recorded to obtain national and comparable Nordic data, as well as data to compare with earlier studies in 2002, 1997, and 1991. Norway had the highest incidence of drug addict deaths by poisoning followed by Denmark, with 8.24 and 6.92 per 100,000 inhabitants, respectively. The death rates in Finland (4.02), Iceland (4.56), and Sweden (3.53) were about half that of Norway and Denmark. Compared with earlier studies, the death rates were unchanged in Denmark and Norway, but increased in Finland, Iceland, and Sweden. In all countries, fewer deaths (29-35%) were recorded in the capital area compared with earlier studies. Females accounted for 11-19% of the fatal poisonings. Iceland deviates with a more equal distribution between men and women (40%). Deaths from methadone overdoses increased in all Nordic countries, and methadone was the main intoxicant in Denmark in 2007, accounting for 51% of the poisonings. In Norway and Sweden, heroin/morphine was still the main intoxicant with a frequency of 68% and 48%, respectively. In Iceland, 3 deaths each were due to heroin/morphine and methadone, respectively. Finland differs from other Nordic countries in having a high number of poisonings caused by buprenorphine and very few caused by heroin/morphine. The total number of buprenorphine deaths in Finland doubled from 16 in 2002 to 32 in 2007, where it constituted 25% of deaths. The general toxicological screening program showed widespread multi-drug use in all countries. The median number of drugs per case varied from 3 to 5. The most frequently detected substances were heroin/morphine, methadone, buprenorphine, tramadol, amphetamine, cocaine, tetrahydrocannabinol, benzodiazepines and ethanol.