MANGANESE DIOXIDE

Synonyms:
Manganese dioxide; Manganese(IV) oxide; 1313-13-9; Manganese peroxide; Manganese superoxide; Manganese oxide (MnO2); Manganese Black; MnO2; Manganese binoxide; Mangandioxid; Black manganese oxide; Cement Black; BOG manganese; Pyrolusite Brown; Bruinsteen [Dutch]; Braunstein [German]; Mangandioxid [German]; Mangaanbioxyde [Dutch]; Mangaandioxyde [Dutch]; C.I. Pigment Brown 8; C.I. Pigment Black 14; MFCD00003463; Manganese (bioxyd de) [French], Manganese (dioxyde de) [French]; Manganese (biossido di) [Italian]; Manganese (diossido di) [Italian]; EINECS 215-202-6; Manganese dioxide, ore of Chapter 26; dioxomanganese; Manganese(IV) oxide, CP; Braunstein; Bruinsteen; Mangaanbioxyde; Mangaandioxyde; Manganese Binoxide Manganese Black; Manganese(IV) oxide, 80-85%, powder; Manganez dioksit; MANGANESE DI OXIDE; MANGANESE DIOXIT; MANGANESE DIOKSIDE; MANGANEZ DIOXIDE; MANGANEZE DİOXİDE; MANGANESE Dİ OXİDE; MANGAN DİOKSİT
 

MANGANESE DIOXIDE
Manganese dioxide

Manganese dioxide
Manganese(IV) oxideMn4O2
Rutile-unit-cell-3D-balls.png
Names
IUPAC names
Manganese oxide
Manganese(IV) oxide
Other names
Pyrolusite, hyperoxide of manganese, black oxide of manganese, manganic oxide
Identifiers
CAS Number    
1313-13-9 check
3D model (JSmol)    
Interactive image
ChEBI    
CHEBI:136511 ☒
ChemSpider    
14117 check
ECHA InfoCard    100.013.821 Edit this at Wikidata
EC Number    
215-202-6
PubChem CID    
14801
RTECS number    
OP0350000
UNII    
TF219GU161 check
CompTox Dashboard (EPA)    
DTXSID6042109 Edit this at Wikidata
InChI[show]
SMILES[show]
Properties
Chemical formula    MnO
2
Molar mass    86.9368 g/mol
Appearance    Brown-black solid
Density    5.026 g/cm3
Melting point    535 °C (995 °F; 808 K) (decomposes)
Solubility in water    insoluble
Magnetic susceptibility (χ)    +2280.0·10−6 cm3/mol[1]
Structure[2]
Crystal structure    Tetragonal, tP6, No. 136
Space group    P42/mnm
Lattice constant    
a = 0.44008 nm, b = 0.44008 nm, c = 0.28745 nm
Formula units (Z)    2
Thermochemistry[3]
Heat capacity (C)    54.1 J·mol−1·K−1
Std molar
entropy (So298)    53.1 J·mol−1·K−1
Std enthalpy of
formation (ΔfH⦵298)    −520.0 kJ·mol−1
Gibbs free energy (ΔfG˚)    −465.1 kJ·mol−1
Hazards
Safety data sheet    ICSC 0175
EU classification (DSD) (outdated)    Harmful (Xn)
Oxidizer (O)
R-phrases (outdated)    R20/22
S-phrases (outdated)    (S2), S25
NFPA 704 (fire diamond)    
NFPA 704 four-colored diamond
112OX
Flash point    535 °C (995 °F; 808 K)
Related compounds
Other anions    Manganese disulfide
Other cations    Technetium dioxide
Rhenium dioxide
Related manganese oxides    Manganese(II) oxide
Manganese(II,III) oxide
Manganese(III) oxide
Manganese heptoxide
Except where otherwise noted, data are given for materials in their standard state (at 25 °C [77 °F], 100 kPa).
☒ verify (what is check☒ ?)
Infobox references
Manganese(IV) oxide is the inorganic compound with the formula MnO
2. This blackish or brown solid occurs naturally as the mineral pyrolusite, which is the main ore of manganese and a component of manganese nodules. The principal use for MnO2 is for dry-cell batteries, such as the alkaline battery and the zinc-carbon battery.[4] MnO2 is also used as a pigment and as a precursor to other manganese compounds, such as KMnO4. It is used as a reagent in organic synthesis, for example, for the oxidation of allylic alcohols. MnO2 in the α polymorph can incorporate a variety of atoms (as well as water molecules) in the "tunnels" or "channels" between the manganese oxide octahedra. There is considerable interest in α-MnO2 as a possible cathode for lithium ion batteries.[5][6]

Structure
Several polymorphs of MnO
2 are claimed, as well as a hydrated form. Like many other dioxides, MnO
2 crystallizes in the rutile crystal structure (this polymorph is called pyrolusite or β-MnO
2), with three-coordinate oxide and octahedral metal centres.[4] MnO
2 is characteristically nonstoichiometric, being deficient in oxygen. The complicated solid-state chemistry of this material is relevant to the lore of "freshly prepared" MnO
2 in organic synthesis.[citation needed] The α-polymorph of MnO
2 has a very open structure with "channels" which can accommodate metal atoms such as silver or barium. α-MnO
2 is often called hollandite, after a closely related mineral.

Production
Naturally occurring manganese dioxide contains impurities and a considerable amount of manganese(III) oxide. Only a limited number of deposits contain the γ modification in purity sufficient for the battery industry.

Production of batteries and ferrite (two of the primary uses of manganese dioxide) requires high purity manganese dioxide. Batteries require "electrolytic manganese dioxide" while ferrites require "chemical manganese dioxide".[7]

Chemical manganese dioxide
One method starts with natural manganese dioxide and converts it using dinitrogen tetroxide and water to a manganese(II) nitrate solution. Evaporation of the water leaves the crystalline nitrate salt. At temperatures of 400 °C, the salt decomposes, releasing N
2O
4 and leaving a residue of purified manganese dioxide.[7] These two steps can be summarized as:

MnO
2 + N
2O
4 ⇌ Mn(NO
3)
2
In another process manganese dioxide is carbothermically reduced to manganese(II) oxide which is dissolved in sulfuric acid. The filtered solution is treated with ammonium carbonate to precipitate MnCO
3. The carbonate is calcined in air to give a mixture of manganese(II) and manganese(IV) oxides. To complete the process, a suspension of this material in sulfuric acid is treated with sodium chlorate. Chloric acid, which forms in situ, converts any Mn(III) and Mn(II) oxides to the dioxide, releasing chlorine as a by-product.[7]

A third process involves manganese heptoxide and manganese monoxide. The two reagents combine with a 1:3 ratio to form manganese dioxide:

Mn
2O
7 + 3 MnO → 5 MnO
2
Lastly, the action of potassium permanganate over manganese sulfate crystals produces the desired oxide.[8]

2 KMnO
4 + 3 MnSO
4 + 2 H
2O→ 5 MnO
2 + K
2SO
4 + 2 H
2SO
4
Electrolytic manganese dioxide
Electrolytic manganese dioxide (EMD) is used in zinc–carbon batteries together with zinc chloride and ammonium chloride. EMD is commonly used in zinc manganese dioxide rechargeable alkaline (Zn RAM) cells also. For these applications, purity is extremely important. EMD is produced in a similar fashion as electrolytic tough pitch (ETP) copper: The manganese dioxide is dissolved in sulfuric acid (sometimes mixed with manganese sulfate) and subjected to a current between two electrodes. The MnO2 dissolves, enters solution as the sulfate, and is deposited on the anode.

Reactions
The important reactions of MnO
2 are associated with its redox, both oxidation and reduction.

Reduction
MnO
2 is the principal precursor to ferromanganese and related alloys, which are widely used in the steel industry. The conversions involve carbothermal reduction using coke:[citation needed]

MnO
2 + 2 C → Mn + 2 CO
The key reactions of MnO
2 in batteries is the one-electron reduction:

MnO
2 + e− + H+
 → MnO(OH)
MnO
2 catalyses several reactions that form O
2. In a classical laboratory demonstration, heating a mixture of potassium chlorate and manganese dioxide produces oxygen gas. Manganese dioxide also catalyses the decomposition of hydrogen peroxide to oxygen and water:

2 H
2O
2 → 2 H
2O + O
2
Manganese dioxide decomposes above about 530 °C to manganese(III) oxide and oxygen. At temperatures close to 1000 °C, the mixed-valence compound Mn
3O
4 forms. Higher temperatures give MnO.

Hot concentrated sulfuric acid reduces the MnO
2 to manganese(II) sulfate:[4]

2 MnO
2 + 2 H
2SO
4 → 2 MnSO
4 + O
2 + 2 H
2O
The reaction of hydrogen chloride with MnO
2 was used by Carl Wilhelm Scheele in the original isolation of chlorine gas in 1774:

MnO
2 + 4 HCl → MnCl
2 + Cl
2 + 2 H
2O
As a source of hydrogen chloride, Scheele treated sodium chloride with concentrated sulfuric acid.[4]

Eo (MnO
2(s) + 4 H+
 + 2 e− ⇌ Mn2+ + 2 H
2O) = +1.23 V
Eo (Cl
2(g) + 2 e− ⇌ 2 Cl−) = +1.36 V
The standard electrode potentials for the half reactions indicate that the reaction is endothermic at pH = 0 (1 M [H+
]), but it is favoured by the lower pH as well as the evolution (and removal) of gaseous chlorine.

This reaction is also a convenient way to remove the manganese dioxide precipitate from the ground glass joints after running a reaction (i. e., an oxidation with potassium permanganate).

Oxidation
Heating a mixture of KOH and MnO
2 in air gives green potassium manganate:

2 MnO
2 + 4 KOH + O
2 → 2 K
2MnO
4 + 2 H
2O
Potassium manganate is the precursor to potassium permanganate, a common oxidant.

Applications
The predominant application of MnO
2 is as a component of dry cell batteries: alkaline batteries and so called Leclanché cell, or zinc–carbon batteries. Approximately 500,000 tonnes are consumed for this application annually.[9] Other industrial applications include the use of MnO
2 as an inorganic pigment in ceramics and in glassmaking.

Organic synthesis
A specialized use of manganese dioxide is as oxidant in organic synthesis.[10] The effectiveness of the reagent depends on the method of preparation, a problem that is typical for other heterogeneous reagents where surface area, among other variables, is a significant factor.[11] The mineral pyrolusite makes a poor reagent. Usually, however, the reagent is generated in situ by treatment of an aqueous solution KMnO
4 with a Mn(II) salt, typically the sulfate. MnO
2 oxidizes allylic alcohols to the corresponding aldehydes or ketones:[12]

cis-RCH=CHCH
2OH + MnO
2 → cis-RCH=CHCHO + MnO + H
2O
The configuration of the double bond is conserved in the reaction. The corresponding acetylenic alcohols are also suitable substrates, although the resulting propargylic aldehydes can be quite reactive. Benzylic and even unactivated alcohols are also good substrates. 1,2-Diols are cleaved by MnO
2 to dialdehydes or diketones. Otherwise, the applications of MnO
2 are numerous, being applicable to many kinds of reactions including amine oxidation, aromatization, oxidative coupling, and thiol oxidation.

See also
List of inorganic pigments

Manganese dioxide is a manganese molecular entity with formula MnO2. It is a manganese molecular entity and a metal oxide.

Molecular Weight of Manganese dioxide:    86.937 g/mol    Computed by PubChem 2.1 (PubChem release 2019.06.18)
Hydrogen Bond Donor Count of Manganese dioxide:    0    Computed by Cactvs 3.4.6.11 (PubChem release 2019.06.18)
Hydrogen Bond Acceptor Count of Manganese dioxide:    2    Computed by Cactvs 3.4.6.11 (PubChem release 2019.06.18)
Rotatable Bond Count of Manganese dioxide:    0    Computed by Cactvs 3.4.6.11 (PubChem release 2019.06.18)
Exact Mass of Manganese dioxide:    86.927872 g/mol    Computed by PubChem 2.1 (PubChem release 2019.06.18)
Monoisotopic Mass of Manganese dioxide:    86.927872 g/mol    Computed by PubChem 2.1 (PubChem release 2019.06.18)
Topological Polar Surface Area of Manganese dioxide:    34.1 Ų    Computed by Cactvs 3.4.6.11 (PubChem release 2019.06.18)
Heavy Atom Count of Manganese dioxide:    3    Computed by PubChem
Formal Charge of Manganese dioxide:    0    Computed by PubChem
Complexity of Manganese dioxide:    18.3    Computed by Cactvs 3.4.6.11 (PubChem release 2019.06.18)
Isotope Atom Count of Manganese dioxide:    0    Computed by PubChem
Defined Atom Stereocenter Count of Manganese dioxide:    0    Computed by PubChem
Undefined Atom Stereocenter Count of Manganese dioxide:    0    Computed by PubChem
Defined Bond Stereocenter Count of Manganese dioxide:    0    Computed by PubChem
Undefined Bond Stereocenter Count of Manganese dioxide:    0    Computed by PubChem
Covalently-Bonded Unit Count of Manganese dioxide:    1    Computed by PubChem
Compound  of Manganese dioxide is Canonicalized    : Yes

sites in France have yielded large numbers of small black blocs. The usual interpretation is that these ‘manganese oxides’ were collected for their colouring properties and used in body decoration, potentially for symbolic expression. Neanderthals habitually used fire and if they needed black material for decoration, soot and charcoal were readily available, whereas obtaining manganese oxides would have incurred considerably higher costs. Compositional analyses lead us to infer that late Neanderthals at Pech-de-l’Azé I were deliberately selecting manganese dioxide. Combustion experiments and thermo-gravimetric measurements demonstrate that manganese dioxide reduces wood’s auto-ignition temperature and substantially increases the rate of char combustion, leading us to conclude that the most beneficial use for manganese dioxide was in fire-making. With archaeological evidence for fire places and the conversion of the manganese dioxide to powder, we argue that Neanderthals at Pech-de-l’Azé I used manganese dioxide in fire-making and produced fire on demand.

Combustion Experiments
Starting from the chemical properties of manganese dioxide, a series of statistically-designed combustion experiments were used to assess whether fire making could be facilitated using wood and either commercial manganese dioxides (coded MD4 to MD6) or powdered material from the Pech-de-l’Azé I blocs (coded MD1 to MD3). Mixtures of wood ‘turnings’ and either manganese dioxide or powdered material from Pech-de-l’Azé I blocs were either heated or contacted with spark-lit tinder; the effects were monitored on video; thermal imaging camera temperature monitoring and XRD of the residues were used in selected cases (Methods).

When heated on their own, the wood turnings released volatiles and produced a small amount of char but neither the volatiles nor the char ignited and no fire resulted (Supplementary Information 3). Similarly, spark-lit tinder did not ignite the wood. By contrast, mixtures of manganese dioxide with wood ignited, both when heated and when in contact with spark-lit tinder. Ignition produced glowing combustion and, in some cases, small red flames; the volatiles did not ignite and no yellow flames were produced (Fig. 2 and Supplementary Information 3). As little as 6% by weight of manganese dioxide MD6 was sufficient to facilitate combustion. Infrared thermal imaging data showed that whilst the wood turnings did not ignite at 350 oC, the mixtures of wood turnings with manganese dioxide could ignite at temperatures from around 250 oC and sustain combustion over a surprisingly wide range of temperatures (Supplementary Information 4). In identical experiments, powdered material from the Pech-de-l’Azé I blocs (MD1, MD2 and MD3) all facilitated the ignition of wood, although one bloc (MD1) was somewhat less effective.
he composition of the black blocs at Pech-de-l’Azé I potentially provides evidence for their probable use. The blocs are predominantly manganese dioxide, not romanèchite and the combustion experiments and TGA have shown that only compositions predominantly containing manganese dioxide would be useful in fire-making. Both manganese dioxide and romanèchite would be useful in decoration32, although whether either would be preferred for decoration over the less ‘costly’ soot or charcoal is debatable. Whether Neanderthals at Pech-de-l’Azé were simply collecting black blocs from one source location or were selecting manganese dioxide in preference to other black materials and from multiple sources is important to our hypothesis that they were deliberately selecting and using manganese dioxide in fire making. Although the quantities and availabilities of different manganese oxides in the Middle Palaeolithic Dordogne region are unknown, there is evidence from both modern sources and from materials collected in the Palaeolithic, for a range of ‘manganese oxide’ materials that were available within reach of Pech-de-l’Azé. Manganese ore outcrops are numerous on the edges of the Massif Central38 and whilst most of the regional manganese ores had been extracted by the early twentieth century32, an original manganese ore source exists in the limestone within a few kilometres of Pech-de-l’Azé. The source contains traces of both manganese dioxide and romanèchite32. Discovery of pyrolusite and romanèchite in a Châtelperronian context at Roc-de-Combe7, thirteen kilometres from Pech-de-l’Azé, also indicates that both materials were available to late Middle Palaeolithic Neanderthals. Pyrolusite, romanèchite, todorokite, hollandite and other black manganese oxide ores were all used in the production of Upper Palaeolithic cave wall images in the vicinity, for example at Lascaux, approximately thirty kilometres from Pech-de-l’Azé19,32,33,34, implying their availability to Palaeolithic foragers.

Without appropriate data on the variation of ‘manganese oxide’ compositions within and between geological sources in the region, the full implications of the Pech-de-l’Azé I bloc compositions for provenance are unknown. Whilst it might be argued that paragenesis might have produced a very variable single source, the relative uniformity of the manganese dioxide content of the blocs contrasts with the between-sample variation in arsenic, barium, cobalt and manganite contents and suggests that the blocs were not collected from one location. Equally, the availability of a range of ‘manganese oxides’ in the region suggests that the blocs were preferentially selected, implying both a capability to recognize the characteristics of these materials - although how this was accomplished is not clear - and an end-use that required the specific properties of manganese dioxide. Pech-de-l’Azé I is not unique and active selection rather than simple collection is supported by the presence of manganese dioxide apparently associated with fire places in the Châtelperronian layers at the Grotte-du-Renne, Arcy-sur-Cure15. The black materials said to be of manganese ores at other Mousterian sites (Supplementary Information 1, Table S1) may provide further evidence when the compositions are published.

Our combustion experiments have shown that manganese dioxide promotes the ignition and combustion of wood and that this is not the case with romanèchite. The Pech-de-l’Azé I blocs would have had to have been ground to powder for use in facilitating fire lighting and there is archaeological evidence for grinding in the form of a grindstone and abraded blocs at Pech-de-l’Azé I27 and at Grotte-du-Renne, Arcy-sur-Cure15. Spark-lit tinder with manganese dioxide powder is one simple yet effective means of starting wood fires with substantially lower wood auto-ignition temperatures and high rate of combustion. Other methods may be envisaged.

The clear benefits for fire-promotion and the presence of manganese dioxide at Neanderthal sites are not evidence that Neanderthals sourced and used manganese dioxide for fire making purposes nor that they did not use the black material for decorative purposes. However, if different ores have similar decorative properties and Neanderthals selected black manganese oxides that have pronounced oxidizing properties compared to others, we might infer that the choices reflect a fire-related end-use and vice-versa. Chalmin32 has shown that specifically for wall ‘painting’, romanèchite produces a more consistent streak than pyrolusite and both are considerably better than manganite; if powdered and dispersed in water, these particular materials are equally effective in decoration. There is apparently no decorative reason for Neanderthals to have favoured manganese oxides over soot and charcoal, or manganese dioxide over other manganese oxides.

In contrast to the “low cost” fire residues, manganese dioxides would have had to have been sourced and transported, at considerably higher costs, which calls for an explanation of such investments outside of body decoration. Our preferred hypothesis is that Neanderthals sourced, selected and transported manganese dioxide for fire making at Pech-de-l’Azé I. Whilst the emphasis here has been on the benefits in fire making, the properties of manganese dioxide could have been exploited in other ways, including improved hafting adhesives16.

It is not suggested that manganese dioxide was necessary for fire making or used by Neanderthals all over their geographical range. How Neanderthals developed the innovation is unclear. In fact, the methods of fire production in the Middle Palaeolithic have not been identified39 and Neanderthals may only have collected fire from wild fires. However, the fact that fire was used as a tool to produce birch-bark pitch already from the early Middle Palaeolithic onward40,41,42 shows that Neanderthals had the capability to control fire from minimally 200,000 years ago. Such a considerable time depth of fire use would be important to a later recognition of the value of manganese dioxide in fire making.

In reviewing the significance of the Female Cosmetic Coalitions (FCC) model in the context of the European Middle Palaeolithic archaeological record, Power, Sommer and Watts8 argue that black “manganese” materials were first present at Pech-de-l’Azé IV and Combe Grenal in the glacial conditions of Marine Isotope Stage (MIS) 4. If analyses shows they are indeed manganese dioxide, these black materials would lend support to an origin in the use of manganese dioxide for fire making in the subsistence challenges of the prolonged cold conditions of MIS 4.

Whilst we can envisage substantial subsistence benefits in the ability to better start, promote and control fire, fire use also comes with a wide range of social benefits and implications43. If Neanderthal engagement with materials and processes held subsistence advantages, it may also have been important in the development of complexity in social relationships. Representing fire promotion by manganese dioxide exclusively as a subsistence benefit, no matter how important, risks understating its possible social and symbolic implications43,44, even though these are notoriously difficult to study in the deep past.

The selection and use of manganese dioxide for fire making is unknown from the ethnographic record of recent hunter gatherers. This unusual behaviour holds potential significance for our understanding of Neanderthal cognitive capabilities through the extent of their knowledge and insights. The actions involved in the preferential selection of a specific, non-combustible material and its use to make fire are not obvious, not intuitive and unlikely to be discovered by repetitive simple trials as might be expected for lithic fracturing, tool forming and tool use. The knowledge and insights suggested by Neanderthal selection of manganese dioxide and use in fire-making are surprising and qualitatively different from the expertise we associate with Neanderthal subsistence patterns from the archaeological record.

We conclude, based on the compositions of the Pech-de-l’Azé I blocs and the availability of different black manganese oxides in the Dordogne region, that Neanderthals were preferentially selecting specifically manganese dioxide blocs. However manganese dioxide does not have clearly evident advantages in decoration over the carbon-rich materials or the other manganese oxides available to Neanderthals. From the combustion and TGA experiments, it is clear that manganese dioxide is an effective facilitator in fire making, reducing the auto-ignition temperature of wood and substantially increasing the rate of combustion. The archaeological evidence of bloc abrasion and grinding stone is consistent with the conversion to powder necessary for use in fire-starting. The intimate association of fire places and manganese dioxide blocs at Pech-de-l’Azé I suggest a use in fire making. We hypothesise that fire-making was manganese dioxide’s most beneficial distinguishing attribute available to Neanderthals. Although we should not exclude the possibility that manganese dioxide was used for decoration and social communication, the combustion, compositional and archaeological strands of evidence lead us to the conclusion that late Neanderthals at Pech-de-l’Azé I were using manganese dioxide in fire-making and by implication were producing fire on demand.

Methods
Materials
Three commercially available manganese dioxide materials were used in the combustion experiments; two reagent grades from Sigma-Aldrich (product reference 310700, coded MD4 and product reference 217646, coded MD6) and a less pure material supplied by Minerals Water Ltd. (coded MD5). A romanèchite, hydrated barium manganese oxide material (coded MD7) from the Schneeberg mine in Saxony, Germany was also used. Its elemental composition is not inconsistent with romanèchite and the XRD-determined structure has close similarities with a romanèchite XRD reference (Supplementary Information 2). This material may not have had precisely the same properties and behaviour as romanèchite material from the Dordogne region.

Three metal oxides were chosen for comparative experiments, all thermally stable oxides, aluminium oxide, zinc oxide and titanium dioxide. All the oxide materials were reagent-grade materials from the Gorlaeus Laboratorium, University of Leiden. Elemental compositions and crystal structures of the manganese oxides are given in Supplementary Information 2.

Three small blackish coloured blocs from the ‘spoil’ of early twentieth century excavations at Pech-de-l’Azé I were studied (coded MD1, MD2 and MD3). These blocs were recovered during the 2004–5 fieldwork season led by M. Soressi; they were in the excavation spoil at the entrance of the cave along with artefacts left by previous excavators, mostly in L. Capitan and D. Peyrony’s 1912 excavation. Two were grey-black pebble-like materials and the third (MD3) had a more slab-like appearance with a reddish colour overlying the grey-black material on one side. Each bloc was examined by optical and scanning electron microscopy (SEM) with EDX and analyzed by XRD and XRF; approximately two grams in total were used in the combustion experiments.

Ten blocs from recorded archaeological contexts in Bordes’ 1970–1 excavations and eleven from Soressi’s 2004–5 excavations were non-destructively analyzed for their XRF compositions and XRD structures. The measured sample set constitutes approximately 5% of the population of blocs when MD1 to MD3 are included. The Bordes’ blocs appeared to have facets or striations suggesting that they had been deliberately abraded. There were no clearly abraded facets on the eleven blocs selected from Soressi’s excavation contexts but there were striations on one bloc. The differences confound two variables, recovery location and apparent use, rendering the interpretation of differences more difficult.

The combustible material was untreated beech wood free from bark, converted into turnings using a hand-held electric drill and 22 mm steel bit. Cotton wool and Ulmus sp. seed were used as tinder materials.

Combustion Experiments
In the combustion experiments, small amounts of the beech wood turnings (1.5 g) or mixtures of beech (1.5 g) with manganese dioxide (0.1 g to 0.5 g) or powdered materials from the Pech-de-l’Azé I blocs or other oxides were placed on a fine steel gauze on a stand within a fume cupboard in a gentle air stream (see Fig. 2). The mixture was heated from below by the flame of a 9.5 cm Sakerhets Tandstickor for fifteen seconds; in some cases the heating time was extended to thirty seconds with a second match. The flame was unable to penetrate the gauze and served to heat the wood via the gauze. For some experiments a Swedish Firesteel 2.0 was used as a source of sparks to light a 0.1 g piece of tinder placed on the surface of the beech turnings. Wherever possible, multiple replication runs were used to validate the outcomes, control runs of beech alone or beech mixed with MD4 or MD6 were used in each phase. In total 120 experimental runs were completed.

The effects were recorded on high definition video. In some experiments the whole combustion process of approximately ten minutes was monitored using either a FLIR A35 or a FLIR T450 thermal imaging camera and combustion temperatures recorded. The temperature data were analyzed using FLIR ResearchIR version 3.4 software (Supplementary Information 4).

Thermo-gravimetric Analysis Methodology
Thermo-gravimetric differential thermal analysis was performed in nitrogen or air atmospheres using a TA-Instruments SDTQ600. A typical sample mass of 12–15 mg was heated to the desired temperature at a ramp rate of 5 °C/min in a total flow rate of 100 ml/min. Beech wood used for the impregnations was ground and sieved to 90 μm. The ground wood (200 mg) was mixed with manganese dioxide to yield 1% by weight, 9% by weight and 23% by weight of manganese dioxide and wood samples. After addition of manganese dioxide the sample was moistened by 1 ml of de-ionized water mixed and oven dried at 60 °C for five hours.