ISSN 10637834, Physics of the Solid State, 2011, Vol. 53, No. 11, pp. 2244–2249. © Pleiades Publishing, Ltd., 2011. Original Russian Text © I.A. Smirnov, B.I. Smirnov, T.S. Orlova, Cz. Sulkovski, H. Misiorek, A. Jezowski, J. Muha, 2011, published in Fizika Tverdogo Tela, 2011, Vol. 53, No. 11, pp. 2133–2137.
Thermopower of Beech Wood Biocarbon I. A. Smirnova, *, B. I. Smirnova, **, T. S. Orlovaa, Cz. Sulkovskib, H. Misiorekb, A. Jezowskib, and J. Muhab a
Ioffe PhysicalTechnical Institute, Russian Academy of Sciences, Politekhnicheskaya ul. 26, St. Petersburg, 194021 Russia * email: [email protected]
** email: [email protected]
b Trzebiatowski Institute of Low Temperature and Structure Research, Polish Academy of Sciences, ul. Okólna 2, 50422 Wroclaw, 50422 Poland Received May 10, 2011
Abstract—This paper reports on measurements of the thermopower S of highporosity samples of beech wood biocarbon with micronsized sap pores aligned with the tree growth direction. The measurements have been performed in the temperature range 5–300 K. The samples have been fabricated by pyrolysis of beech wood in an argon flow at different carbonization temperatures (Tcarb). The thermopower S has been measured both along and across the sap pores, thus offering a possibility of assessing its anisotropy. The curves S(Tcarb) have revealed a noticeable increase of S for Tcarb < 1000°C for all the measurement temperatures. This finding fits to the published data obtained for other physical parameters, including the electrical conductivity of these biocarbons, which suggests that for Tcarb ~ 1000°C they undergo a phase transition of the insulator(at Tcarb < 1000°C)–metal(at Tcarb> 1000°C) type. The existence of this transition is attested also by the character of the temperature dependences S(T) of beech wood biocarbon samples prepared at Tcarb above and below 1000°C. DOI: 10.1134/S1063783411110291
1. INTRODUCTION Investigation of the physical properties of carbon materials has been attracting extensive experimental and theoretical interest for a long time. Publications reporting on studies of their thermopower S are, how ever, fairly scanty, although the relevant data could provide valuable information bearing on the band structure parameters of the materials in question, as well as on the nature of the electron–phonon interac tions involved. The first comprehensive review addressing data available in the literature on the thermopower of car bon materials was published in 1916 . In the subse quent years, no adequate theoretical and experimental activities in the field of the thermopower of these materials have been witnessed, although systematic studies of the temperature dependences S(T) of the numerous representatives making up the family of car bon materials have been carried out and continue to be done. Recent publications reveal interest in the inves tigation of the dependences S(T) observed in new car bon materials, more specifically, graphene , nanop orous carbon , and biocarbon [4, 5]. The present report deals with a study of S(T) of the beech wood biocarbon. Biocarbon materials are prepared by pyrolysis (car bonization) of the corresponding tree wood in an argon flow carried out at various carbonization tem peratures Tcarb [6–9]. The biocarbon fabricated in this
way is known under different names—precursor, car bon template, carbon preform. One has to distinguish between two major kinds of biocarbon prepared from tree wood. Biocarbon of the first kind is fabricated by carbonization of natural tree wood. The porosity of this biocarbon provided by sap channels may reach as high as ~75 vol %. The empty channel pores (micronsized sap channels) are aligned with the tree growth direction. This accounts for the anisotropy in the physical parameters measured along and across the sap pores of a sample. Biocarbon of the second kind encompasses materi als fabricated by pressing sawdust (“microwood fiber”) of the corresponding tree wood species [10, 11] which, like biocarbon of the first kind, were subjected to car bonization. This carbon material is usually referred to as “wood artificial fiberboards” (WAF). While the physical parameters of these materials do not reveal anisotropy, it may materialize under conditions in which pressing produces an oriented arrangement among the fiberboard particles with intact sap chan nels. We are going to dwell now on some physical param eters of biocarbon which will be needed in the discus sion of the results of the present study to follow. (1) An Xray diffraction investigation of a number of tree wood biocarbons [12–15] has culminated in setting up a model realizing the structure of their car bon templates. It was demonstrated that these tem
THERMOPOWER OF BEECH WOOD BIOCARBON
PHYSICS OF THE SOLID STATE
12 2 S, μV/K
plates form from nanocomposites  made up of amorphous carbon and nanocrystallites of two types, more specifically, threedimensional “graphite frag ments” composed, depending on the actual magni tude of Tcarb, of two and more carbon layers, and two dimensional layers of graphene. As Tcarb increases, the fraction of the amorphous component in nanocom posites decreases as compared to that of the nanocrys talline phase. (2) A study of the impact exerted by annealing tem perature THT on various kinds of soft and hard carbon and by the Tcarb for biocarbons of both kinds on the behavior of their physicochemical, structural, mechanical, acoustic, thermal and electrical proper ties [8, 10, 11, 16–18] revealed that at the “critical” temperature Tcr corresponding to Tcarb(THT) ~ 1000°C these materials undergo an insulator (at Tcarb(THT) < 1000°C) to metal (at Tcarb(THT) > 1000°C)type phase transition. This effect became particularly man ifest in the WAF and beech wood biocarbon in the behavior of the electrical resistivity ρ(Tcarb) [11, 16]. This suggested that for samples with Tcarb < 1000°C and Tcarb > 1000°C the behavior of ρ(Tcarb) is mediated primarily by the amorphous and nanocrystalline phases of the carbon nanocomposite, respectively. For samples with Tcarb < 1000°C the ρ(T) dependence fits Mott’s law for variablerange hopping conduction in systems with exponential distribution of the density of localized states near the Fermi level, while samples fabricated at Tcarb > 1000°C should be assigned to dis ordered metallic systems with band conduction . The literature data on thermopower in biocarbons of both types are fairly scarce. Only two papers report measurement of S within the 5–300K interval of sapele wood preforms prepared at Tcarb = 1000°C  and white pine preforms fabricated at Tcarb = 1000 and 2400°C . There are no publications on measure ment of S for the WAF biocarbon. Our major goal was to broaden the scope of biocar bon materials covered by S(T) studies. The program assumed measurements of S(T) within the 5–300K interval on beech biocarbon samples prepared at Tcarb =800, 1000, and 2400°C. A theoretical analysis  showed thermopower to be very sensitive to the mechanism of carrier conduc tion in the material under study. This manifests itself both in the temperature behavior of thermopower and in its magnitude. In the hopping conduction region at low temperatures S ~ T 0.5, and it is larger than that in a disordered bandconduction metallic system. This brings us to the second problem to be solved in the present study, more specifically, to an analysis of the available experimental S(T) data on the beech bio carbon samples prepared at Tcarb below (800°C) and above (1000, 2400°C) the critical temperature Tcr of the insulatormetal phase transition observed in this system [11, 16], with a subsequent comparison of the
T, K Fig. 1. Thermopower of the BEC800 sample plotted vs. temperature: (1) S|| and (2) S⊥.
results obtained with the theoretical conclusions drawn in . 2. SAMPLES AND EXPERIMENTAL TECHNIQUE The thermopower of beech biocarbon was mea sured on samples fabricated at Tcarb = 800 (BEC 800), 1000 (BEC1000), and 2400°C (BEC2400) which were used by us in our preceding studies [15, 16] of their electrical and thermal conductivities. The samples were prepared by the standard technique employed in [6–9] and mentioned briefly in Introduc tion. More detailed information on these samples (including the relevant Xray diffraction data) can be found in [15, 16]. We are going to cite here only those of them which may be of immediate interest for the present study. The average density of carbon templates of all the beech biocarbon samples studied is ~1.49 g/cm3. The porosity of the BEC800 samples is ~47 vol %, and that of the BEC1000 and BEC 2400, ~60 vol %. The thermopower was measured in the 5–300K interval by the standard diffusion tech nique. 3. EXPERIMENTAL RESULTS AND DISCUSSION Figures 1 and 2 plot experimental data on the behavior with temperature of the thermopower mea sured along (S||) and across (S⊥) the empty sap chan nels of the BEC800, BEC1000, and BEC400 samples. Presented below are our comments and con clusions concerning the results obtained. 2011
SMIRNOV et al. 3 6
1 2 βS, β'S
2 3 4 S, μV/K
200 T, K
Fig. 3. (1, 2) Thermopower anisotropy βS of the (1) BE
C800 and (2) BEC1000 samples and (3) β 'S of the BEC2400 sample. The dashdotted straight line cor responds to βS = 1.
for the BEC800 and BEC1000 samples (~2.5– 2.0) and approaches 1 for BEC2400.
T, K Fig. 2. Temperature dependences of the thermopower of the BEC1000 ((1) S||and (2) S⊥)) and BEC2400 ((3) S|| and (4) S⊥).
(1) The thermopower of the BEC800 and BEC 2400 samples is positive throughout the temperature interval covered, whereas that of the BEC1000 sam ples reverses sign (from “plus” to “minus”) at temper atures below 65–75 K. (2)The behavior of S(T) of beech biocarbons with temperature follows a pattern identical with that of the sapele  and white pine  preforms. (3) Within the 100–300K interval, all beech bio carbon samples exhibit thermopower anisotropy (Fig. 3). In the BEC1000 and BEC2400 samples, however, it close to disappears below 50–75 K (Fig. 2). This does not occur with the BEC800 sample. It appears pertinent to point out here two more fea tures in the beech biocarbon thermopower anisotropy observed in the 100–300K interval. (a) The anisotropy manifest in the BEC800 and BEC1000 samples can be characterized by the ratio βS = S⊥/S||, as was the case with the sapele biocarbon . For the BEC2400 sample, however, β S' = S||/S⊥. (b) For the thermal conductivity anisotropies βκ = κ||/κ⊥)  measured on BEC800, BEC1000, and BEC2400 samples in the temperature range 100 300K temperature are confined to the 2–2.6 interval and do not depend noticeably on the magnitude of Tcarb. In the case of thermopower, the magnitude of the anisotropy depends substantially on Tcarb. It is larger
(4) The data visualized in Figs. 4 and 5 may be classed among the most interesting S(T) results obtained for beech biocarbon. Figure 4 plots the data for S||(T) amassed for beech biocarbon samples vs. Tcarb. They are confronted by the results obtained  for S(THT) (soft carbon) annealed at different temper atures THT. The inset to Fig. 4 visualizes the values of S⊥(Tcarb) for all the beech biocarbon samples studied in the work (Figs. 1 and 2).
a b c
12 8 4 0 500
1000 2000 3000 Tcarb, °C
1500 2000 Tcarb(THT), °C
Fig. 4. Behavior of the thermopower of (1, 2, 4) beech bio carbon for different Tcarb and (3) soft carbon  plotted vs. annealing temperature THT: (1, 2) S|| for 300 and 100 K, respectively; (4) S⊥ for 300 K. (a–c) Data for the BEC 800, BEC1000, and BEC2400 samples, respectively. The magnitude of S at 300 K for (3) soft carbon is refer enced to platinum.
PHYSICS OF THE SOLID STATE
THERMOPOWER OF BEECH WOOD BIOCARBON
2 S, μV/K
3 T0.5 1
100 T, K
Fig. 5. Log plots of the temperature dependences of ther mopower S|| for (1) BEC800 and (2) BEC2400 sam ples.
As seen from the S||(Tcarb) and S⊥(Tcarb) curves pre sented in Fig. 4 for beech biocarbon, for Tcarb < 1000°C considered to be the critical temperature (Tcr) for the insulator–metal phase transition in these materials , S demonstrates a noticeable growth in magnitude. This suggests a conclusion that the data obtained for the thermopower of beech biocarbon may be regarded as evidence in it of the insulator–metal phase transition. Turning now to the presence or absence for the beech biocarbon of a maximum in the S(Tcarb) curve, which was observed to exist in annealed samples of soft carbon , no straightforward con clusion can be drawn at present on this issue, because we have not carried out measurements of S(Tcarb) on samples of beech biocarbon subjected to carboniza tion at Tcarb in the 1500–2000°C interval. (5) Figure 5 plots on a log scale the S||(T) data for the BEC800 and BEC2400 samples. For the first sample, in the 5–50K lowtemperature region S|| ~ T 0.5, while for T > 50 K, S|| ~ T. For the second sample, PHYSICS OF THE SOLID STATE
E 200 T, K
Fig. 6. Temperature dependences of the thermopower of pyrolytic graphite A, C, D, E and of a compact graphite sample B (see ). For samples A, B, and C, the measure ments of S were performed along the c axis, and for sam ples D and E, along the a axis. Dashed lines 1–3 visualize beech biocarbon data (samples BEC800, BEC2400, and BEC1000, respectively). The dashdotted straight line is drawn for S = 0.
S|| ~ T throughout the temperature range covered (5– 300 K). The temperature dependences of the S||(T) thermopower and its magnitude fit well to the theoret ical conclusions drawn in  for the behavior of the S(T) dependence of materials in the hopping and metallic conduction regions. Summing up the above measurements of S(T), one can say that what we observe in beech biocarbon samples prepared at Tcarb < 1000°C is a manifestation of the hopping, and, at Tcarb > 1000°C, that of metallic carrier conduction. We consider a few points bearing on the carbon templates of biocarbon prepared from different tree species. As mentioned in Introduction, as well as in , these carbon templates originate from specific carbon nanocomposites composed of amorphous car bon, “graphite fragments,” and graphene. It appeared 2011
SMIRNOV et al.
instructive to compare the thermopowers generated by bulk materials making up the carbon nanocomposites with S of the beech biocarbon studied. Thermopower of amorphous carbon, shungite , is ~15.5 and 5.5 μV/K for 300 and 100 K, respec tively, i.e., noticeably in excess of the values of S for the BEC800 sample, as this should be, if we accept the tentative structure of biocarbon proposed in this study. Figure 6 summarizes the data for S(T) of pyrolytic and compact graphite measured along the a and c axes at 100–300 K . Also shown are the S(T) graphs obtained in the present work for the BEC800, BE C1000, and BEC2400 samples. The results obtained turned out instructive and quite unexpected. The thermopower of all the beech biocarbon samples was found to coincide with the data amassed for S(T) of graphite measured along the c axis, both in sign and magnitude and the temperature behavior. The thermopower coefficient of graphene mea sured in the 4–380K interval  was found to differ strongly both in magnitude (from a few units at 4 K to ~100 μV/K at 300 K) and in sign (it is negative) from the behavior of S(T) of beech biocarbon. A numerical comparison of the values of S(T) mea sured on the materials making up the carbon nanocom posite with corresponding data for beech biocarbon is anything but a simple problem. One would have to take into account the percentage of the original components making up the nanocomposite, which actually would not present any difficulties, but what is really hard, is to assess the effect produced on the magnitude of S by materials of which the nanocomposite is composed as their size drops to a few nanometers [23, 24]. 4. CONCLUSIONS The paper presented the first measurements of the thermopower S in the 5–300K interval of beech bio carbon samples prepared by pyrolysis of beech tree wood in an argon flow at carbonization temperatures Tcarb = 800, 1000, and 2400°C. The samples studied had a high sap channel porosity (46–60 vol %). The micronsized empty channels in the samples were aligned with the tree growth direction. The samples used in the S(T) study were the same on which struc tural parameters, electrical resistivity and thermal conductivity were earlier measured [15, 16]. The measurements of S(T) and of the dependences of S on the magnitude of Tcarb at which the beech bio carbon samples were prepared revealed the existence in them of an insulator (for Tcarb < 1000°C)–metal (for Tcarb > 1000°C) phase transition at Tcarb ~ 1000°C. The temperature dependences S(T) measured on samples prepared at Tcarb = 800 and 2400°C are con sistent with the available theoretical picture  in which the first of the above samples are dominated by hopping, and the second, by metallic carrier conduc tion.
The anisotropy of S was estimated for all the sam ples investigated. ACKNOWLEDGMENTS We are grateful to K.T. Faber (Northwestern Uni versity, United States) and J. MartinezFernandez (Universidad de Sevilla, Spain) for providing the beech wood biocarbon samples for the present study. The study was supported by the Presidium of the Russian Academy of Sciences (program P03), the Ministry of Science and Higher Education of Poland (project no. 202259939), and the Bilateral Agreement between the Polish and Russian Academies of Sci ences. REFERENCES 1. M. La Rosa, Nuovo Cimento 12, 284 (1916). 2. J. H. Seol, I. Jo, A. L. Moore, L. Lindsay, Z. H. Aitken, M. T. Petter, X. Li, Z. Yao, R. Huang, D. Briodo, N. Mingo, R. S. Ruoff, and L. Shi, Science (Washing ton) 328, 213 (2010). 3. V. V. Popov, S. K. Gordeev, A. V. Grechinskaya, and A. M. Danishevskii, Phys. Solid State 44 (4), 789 (2002). 4. L. S. Parfen’eva, B. I. Smirnov, I. A. Smirnov, D. Wlosewicz, H. Misiorek, A. Jezowski, A. R. de Arel lanoLopez, and J. MartinezFernandez, Phys. Solid State 51 (11), 2252 (2009). 5. A. T. Burkov, S. V. Novikov, B. I. Smirnov, I. A. Smirnov, Cz. Sulkowski, and A. Jezowski, Phys. Solid State 52 (11), 2333 (2010). 6. A. R. de ArellanoLopez, J. MartinezFernandez, P. Gonzalez, D. DomíngezRodriguez, V. Fernandez Quero, and M. Singh, Int. J. Appl. Ceram. Technol. 1, 56 (2004). 7. P. Greil, T. Lifka, and A. Kaindl, J. Eur. Ceram. Soc. 18, 1961 (1998). 8. C. E. Byrne and D. C. Nagle, Carbon 35, 267 (1997). 9. C. Zollfrank and H. Siber, J. Eur. Ceram. Soc. 24, 495 (2004). 10. A. K. Kercher and D. C. Nagle, Carbon 41, 15 (2003). 11. V. V. Popov, T. S. Orlova, E. Enrique Magarino, V. A. Bautista, and J. MartinezFernandez, Phys. Solid State 53 (2), 276 (2011). 12. L. S. Parfen’eva, T. S. Orlova, N. F. Kartenko, N. V. Sha renkova, B. I. Smirnov, I. A. Smirnov, H. Misiorek, A. Jezowski, J. Mucha, A. R. de ArellanoLopez, J. MartinezFernandez, and F. M. VarelaFeria, Phys. Solid State 48 (3), 441 (2006). 13. L. S. Parfen’eva, T. S. Orlova, N. F. Kartenko, N. V. Sharenkova, B. I. Smirnov, I. A. Smirnov, H. Misiorek, A. Jezowski, T. E. Wilkes, and K. T. Faber, Phys. Solid State 50 (12), 2245 (2008). 14. L. S. Parfen’eva, T. S. Orlova, N. F. Kartenko, N. V. Sharenkova, B. I. Smirnov, I. A. Smirnov, H. Misiorek, A. Jezowski, J. Mucha, A. R. de Arellano Lopez, and J. MartinezFernandez, Phys. Solid State 51 (12), 2023 (2009).
PHYSICS OF THE SOLID STATE
THERMOPOWER OF BEECH WOOD BIOCARBON 15. L. S. Parfen’eva, T. S. Orlova, N. F. Kartenko, N. V. Sharenkova, B. I. Smirnov, I. A. Smirnov, H. Misiorek, A. Jezowski, T. E. Wilkes, and K. T. Faber, Phys. Solid State 52 (6), 1115 (2010). 16. L. S. Parfen’eva, T. S. Orlova, N. F. Kartenko, B. I. Smirnov, I. A. Smirnov, H. Misiorek, A. Jezowski, J. Muha, and M. C. Vera, Phys. Solid State 53 (11), 2938 (2011). 17. J. C. Hernandez, I. HernandezCalderon, C. A. Luengo, and R. Tsu, Carbon 20, 201 (1982). 18. F. Carmona, P, Delhaes, G. Keryer, and J. P. Manceau, Solid State Commun. 14, 1183 (1974). 19. I. P. Zvyagin, Phys. Status Solidi B 58, 443 (1973).
PHYSICS OF THE SOLID STATE
20. E. E. Loebner, Phys. Rev. 102, 46 (1956). 21. L. S. Parfeneva, T. I. Volkonskaya, V. V. Tikhonov, I. N. Kulikova, I. A. Smirnov, N. N. Rozhkova, and A. Z. Zaidenberg, Phys. Solid State 36 (4), 625 (1994). 22. L. C. F. Blackman, P. H. Dundas, and A. R. Ubbelohde, Proc. R. Soc. London, Ser. A 255, 293 (1960). 23. S. V. Faleev and F. Leonard, Phys. Rev. B: Condens. Matter 77, 214304 (2008). 24. A. V. Dmitriev and I. P. Zvyagin, Phys.–Usp. 53 (8), 805 (2010).
Translated by G. Skrebtsov