NANO LETTERS
Nanoscale Thermal Analysis of an Energetic Material
2006 Vol. 6, No. 9 2145-2149
William P. King,* Shubham Saxena, and Brent A. Nelson Woodruff School of Mechanical Engineering, Georgia Insitute of Technology, Atlanta, Georgia 30080
Brandon L. Weeks and Rajasekar Pitchimani Department of Chemical Engineering, Texas Tech UniVersity, Lubbock, Texas 79409 Received May 25, 2006; Revised Manuscript Received August 4, 2006
This paper reports nanoscale thermal analysis on a polycrystalline energetic material using a heated atomic force microscope cantilever tip. The heated tip performs highly local melting, evaporation, and decomposition with modifications in the material from 100 nm to several micrometers. The local thermochemical reactions have strong temperature dependence, and the thermal analysis includes measurements of the temperature-dependent material response over the range of 25-500 °C. The measurements interrogate energetic material thermomechanical behavior at unprecedented length scales, and as such the techniques could be used to provide new insight into the fundamental behavior of energetic materials. Energetic materials are materials that exhibit a dramatic release of stored chemical energy as thermal and mechanical energies. The primary difference between an energetic material and any material that undergoes a chemical decomposition process is the rate at which the decomposition occurs. The decomposition rate is determined by a number of factors including the particle characteristics (chemical composition, size, morphology), the magnitude and duration of the stimulus that triggered the chemical reactions, and the confinement of the energetic material. For explosives, the rate and amount of energy released is normally sufficient to establish a self-sustaining shock known as detonation. Energetic materials often have nanometer-scale polycrystallinity, voids, and/or defects, and it is widely believed that nanoscale properties and phenomena within these materials play a key role in their macroscopic behavior.1-3 One example of nanometer-scale phenomena in energetic materials is “hot spots,” which are nanoscale to microscale voids within the energetic material, which play a key role in energetic material decomposition.4 When exposed to an initiation stimulus, these hot spots act as ignition sites that grow in temperature, size, and pressure, leading to a deflagration or detonation. The formation of voids within an energetic material is not easily controllable during materials synthesis, but has a dramatic impact on the * Corresponding author. E-mail:
[email protected]. Phone: (404) 385-4224. 10.1021/nl061196p CCC: $33.50 Published on Web 08/29/2006
© 2006 American Chemical Society
sensitivity and performance of the energetic material. A deep understanding of the formation and behavior of voids within the energetic material would enable an understanding of detonation processes on the nanoscale level. The hot spots are but one of several important nanoscale thermomechanical properties of energetic materials, none of which have been studied extensively because of the lack of appropriate experimental probes. Nanodectonics techniques, such as those reported in this paper, could enable improved design of energetic materials and ultimately yield safer and more powerful explosives. For measuring, sensing, and manipulating at the nanometer scale, the atomic force microscope (AFM) has become one of the most widely used instruments.5 AFM can detect and form nanometer-scale features on surfaces. In the case of energetic materials, AFM has been used to investigate nanoscale defects, plastic flow, and failures on cyclotrimethylenetrinitramine (RDX) crystalline energetic material.6 AFM has also been used to measure kinetic processes such as nanoscale phase transitions elucidating structures not observed in macroscale experiements.7 These studies have shown that nanoscale investigations are important for fundamental understanding of mechanical properties, combustion/detonation characteristics, and sensitivity of energetic materials. Although it is known that the behavior of energetic materials depends critically upon temperature and heat flow at the nano- to microscale levels, to date, no published study has used a nanoscale probe to investigate these phenomena directly. This paper uses a heated atomic force microscope cantilever tip to perform nanoscale thermal measurements on a polycrystalline energetic material. The nanoscale thermal analysis was performed using AFM cantilevers having integrated heater-thermometers fabricated in our group at the Georgia Insitute of Technology. Figure 1 shows one of these heatable AFM cantilevers, which is similar to heated AFM cantilevers developed for data storage.8-10 This type of cantileverhasalsobeenshowntobeusefulfornanomanufacturing,11-13 metrology,14 and nanoscale thermophysical measurements.15,16 The cantilevers were made of doped single-crystal silicon
Figure 1. Top: silicon heater-cantilever fabricated at the Georgia Insitute of Technology. Bottom: method of nanoscale thermal decomposition with one of the heated probe tips.
Figure 2. Lithograpic marks written into the PETN film using the heated tip. Little to no pileup or other residue is apparent on the PETN surface, indicating that the PETN was decomposed.
and had tips with a radius of curvature near 20 nm. In these cantilevers, both heating and thermometry occurs in the solidstate resistor at the base of the tip. The instrumentation, temperature calibration, and operation of these probes has been described elsewhere8-16 and not reviewed here. Previously, micrometer-scale thermal analysis using heated AFM probes made from fine metal wire was demonstrated17 in which the probe measured heat flow from the tip to a substrate for the purpose of measuring temperature-dependent phase transitions. This paper achieves nanometer-scale thermal measurements using silicon probes. The present work achieves significantly higher temperatures and finer resolution than has been demonstrated previously for microthermal analysis. The energetic material used was pentaerythritol tetranitrate (PETN), which was selected because it is one of the most commonly used secondary explosives and contains only one polymorph at room temperature. Figure 1 shows the basic experimental configuration. A thin film of PETN was prepared at a thickness of 250 nm on a glass slide. The PETN powders (synthesized by Mound EG&G and supplied by Lawrence Livermore National Laboratory) were dissolved in acetone to prepare a 0.2 M solution. This solution was spin coated on to a glass substrate at room temperature using a spin processor at 2500 rpm for 40 s. The films were dried in air at room temperature for 30 min and were characterized by noncontact AFM, giving a film thickness of ∼250 nm. When the heated AFM cantilever tip was scanned over the energetic material, heating from the tip could induce nanoscale melting and/or decomposition in the energetic material. It was possible to perform metrology of the energetic material using a cold tip, both before and after thermal writing. Local thermal decomposition with a heated probe tip provides a unique method of controlling both the size and spatial resolution of voids in the energetic material. In general, the voids within an energetic material depend on the particle size and shape and the overall density of the
final product. The ability to tailor synthetic voids would enable new ways of controlling energetic phenomena. Figure 2 show a simple “+” pattern written in the PETN film, demonstrating the high spatial resolution and registry of the technique. For each of the two lines of the “+”, the cantilever was at 215 °C and scanned at 0.1 Hz for 60 s. The tip load force was 7 nN. The depth of the feature was 300 nm, which closely matched the thickness of the film. There was no noticeable pileup or residue, indicating that the material was completely decomposed or evaporated during the thermal writing. Figure 3 shows the effect of tip temperature on the energetic material response. In this experiment, the heated tip raster scanned lines at five different temperatures. All of the lines were written at 0.1 Hz for 60 s and a tip load force of 50 nN. The lowest temperature tested, 54 °C, produced no lithographic mark on the PETN. However, at 99 °C and above the heated tip was able to write into the PETN; the region of PETN reaction was wider for increasing temperature. The increased reaction area may have been due to increased heating from the tip, or by diffusion of a thermomechanical reaction in the PETN film. For the areas decomposed at higher temperatures, the PETN crystals near the decomposed area were noticeably larger than those in the unmodified sample regions, suggesting that this type of measurement may be useful for studying grain coarsening and aging in energetic materials.18 To test the rate of material reaction, several lines were written at 215 °C over several scan speeds from 0.1 to 2 Hz, showing repeatable conditions but no difference in the reacted region due to tip speed. However, it is possible that scan speed modulates the reaction for a tip speed outside of this range. A second experiment tested the rate of material reacted by scanning the heated tip over a 5 µm square of the PETN film. As shown in Figure 4, the tip was raster scanned over the sample in one continuous path, making only one pass over the sample in the slow scan direction. In the images of Figure 4, the slow scan began at the “south” end of the image
2146
Nano Lett., Vol. 6, No. 9, 2006
Figure 3. The heated tip was scanned over the PETN in five lines, each at a different temperature. This image shows the PETN after all five experiments. The decomposed PETN is much wider for the writing at higher temperatures, possibly due to the diffusion of heat or the thermochemical reaction through the PETN. Additionally, the crystal grains are much larger in the vicinity of higher heating.
and moved “north” in only one pass such that the tip did not scan over the same region twice. For these experiments, the cantilever was heated to 215 °C and scanned across the width of the sample at 0.1 Hz with a tip load force of 47 nN. For the first experiment, the heated tip scanned over the sample in 1290 s. In the postreaction metrology of Figure 4, much of the PETN that was heated was removed, but unlike the decomposed lines of Figures 2 and 3, some of the PETN filled in behind. The width of the region corresponded to the 5 µm scan during writing, but the length of the decomposed region was somewhat less, with PETN present at the location where the scan began. Furthermore, it appears as if the polycrystalline structure of the PETN orients in a columnar fashion in the north-south direction in Figure 4. A second 5 µm square was written on a fresh area of PETN, under identical conditions, except for an increased scan speed that resulted in a total scan time of 660 s. For this second, faster experiment, significantly less PETN was removed and the columnar grain structure of the PETN is even more apparent. When heated, the PETN can either go through a phase transition (sublimation or melt/ evaporation) into the gas phase or decompose. We hypothesize that the PETN was melted or evaporated at the heated tip, and subsequently recondensed onto the previously Nano Lett., Vol. 6, No. 9, 2006
Figure 4. Top: illustration showing the path of the heated probe tip in which the tip scanned through the area of a square exactly one time. Middle and bottom: images of the decomposed PETN following two different experiments. The slower scan decomposed a larger area than the faster scan. In both cases, the PETN either diffused or condensed to form columnar features.
scanned area. However, not all of the material was recondensed, which suggests that some of the PETN may have decomposed. The recondensed PETN is mostly in the south of the region where the tip scan began. We hypothesize that the PETN condensed only in the south end of the scan because the north end was heated last, leaving a temperature gradient. We hypothesize that the high temperature of the tip drove the liquid or vapor PETN away from the tip, which resulted in PETN condensed on the southern end of the scan only, which was cooler. The condensed PETN formed columnar structures that generally lie in the north-south direction, which is a behavior that is consistent with the temperature gradient being strongest in the north-south direction. Less material condensed within the scanned square for the longer scan and slower tip speed. The longer dwell time of the heated tip may have allowed the melted/ 2147
evaporated PETN to diffuse farther from the heated source. This technique for manipulating the micro/nanostructure of polycrystalline energetic materials could be used to study phenomena such as diffusion rates and produce controlled nanoscale features of arbitrary shape and spacing to investigate propagation between voids and/or oriented crystallites. Finally, the PETN was studied using the heated cantilever tip for measuring the threshold decomposition temperature of the film. This nanothermal analysis (nTA) is similar to a well-known microscopy technique known as microthermal analysis.17,19 By measuring both heat flow and temperature, it is possible to measure thermophysical properties of a material surface. The nTA was performed by holding the heated silicon cantilever tip in contact with the single-crystal PETN and heating the probe slowly. The heating was a slow DC ramp superimposed by a small AC dither. A lock-in amplifier measured the voltage drop across the cantilever to track the phase lag between the supplied AC voltage and the corresponding cantilever temperature response. The cantilever power and the phase lag were recorded at each DC temperature point once the phase lag stabilized, usually after 30-60 s. Figure 5 shows the results of the nanothermal analysis experiment. The tip completely penetrated into the PETN surface, and so the topography image of Figure 5 shows a relatively large indent that is approximately the size of the tip. Figure 5 shows cantilever power and temperature phase lag as a function of temperature, as well as the temperature derivative of power and phase lag. Both images show a discontinuous jump at approximately 61 °C, which corresponded to the formation of the indent shown. The phase lag and supplied power to the cantilever depend on the impedance to heat flow between the cantilever and its environment. Changes to the thermal impedance due to phase transitions in the substrate affect the phase and power signals of the cantilever. Although discontinuities in the specific heat that occur at phase transitions can affect the thermal impedance, the relatively long time scale over which the individual data points are acquired minimizes the contribution this may have to the observed data.20 The primary contribution to thermal impedance change comes from the change in thermal contact between the cantilever tip and PETN. When the PETN melts or decomposes, the tip sinks into the substrate, improving heat transfer both from the tip to the substrate and also from the cantilever legs to the substrate because of the decreased separation distance. For our heated cantilevers, the primary effect comes from the enhancement of heat transfer from the heater legs to the substrate.9 If the tip penetrates into the substrate, which occurs when the phase change temperature is reached, then the corresponding change in the thermal resistance between the cantilever and the surface yields a measurable change in the power and phase response of the heated cantilever. The measured decomposition temperature of 61 °C is lower than the bulk value of 144 °C. The decomposition temperature was not a function of pressure for load pressures in the range 0.3-3 GPa. The observation of melting temperature depression was observed in 77 experiments on 4 individual PETN samples. 2148
Figure 5. Results of the nanothermal analysis experiment. Top: single indent formed for the nTA experiment. Middle and bottom: power and phase lag of the probe signal as a function of temperature. The discontinuity in the power and phase lag signals at 61 °C correspond to the indent formation.
We hypothesize that the observed melting temperature depression was due to a size effect in the PETN phase change properties, a phenomena that is well documented in material science.21 This paper presents new methods for testing the nanometer-scale thermomechanochemical response of an energetic material. Thermochemical reactions can be induced on the thin film materials by controlling the temperature of the probe. The experiments investigate propagation of the thermochemical reaction based on size, shape, spacing, and anisotropy. The paper also presents nanothermal analysis performed on PETN in which phase transitions can be observed Nano Lett., Vol. 6, No. 9, 2006
at a much lower temperature than the bulk. This technique could be used to investigate thermophysical phenomena in any crystalling or polycrystalline material. The ability to manipulate the micro/nanostructure of polycrystalline materials could be used to study phenomena such as diffusion rates, phase transitions, and perform lithography in a wide variety of nanomaterials beyond energetic materials. Acknowledgment. This work was supported by an NSF CAREER Award for W.P.K. (CTS-0338888) and by the Office of Naval Research. References (1) Armstrong, R. W. J. Phys. IV 1995, 5, 89-102. (2) Dlott, D. D. Mater. Sci. Technol. 2006, 22, 463-473. (3) Armstrong, R. W.; Baschung, B.; Booth, D. W.; Samirant, M. Nano Lett. 2003, 3, 253-255. (4) Tarver, C. M.; Chidester, S. K.; Nichols, A. L. J. Phys. Chem. 1996, 100, 5794-5799. (5) Binnig, G.; Quate, C. F.; Gerber, C. Phys. ReV. Lett. 1986, 56, 930933. (6) Sharma, J.; Armstrong, R. W.; Elban, W. L.; Coffey, C. S.; Sandusky, H. W. Appl. Phys. Lett. 2001, 78, 457-459. (7) Weeks, B. L.; Ruddle, C. M.; Zaug, J. M.; Cook, D. J. Ultramicroscopy 2002, 93, 19-23.
Nano Lett., Vol. 6, No. 9, 2006
(8) Chui, B. W.; Stowe, T. D.; Ju, Y. S.; Goodson, K. E.; Kenny, T. W.; Mamin, H. J.; Terris, B. D.; Ried, R. P. J. Microelectromech. Syst. 1998, 7, 69-78. (9) King, W. P.; Kenny, T. W.; Goodson, K. E.; Cross, G.; Despont, M.; Durig, U.; Rothuizen, H.; Binnig, G. K.; Vettiger, P. Appl. Phys. Lett. 2001, 78, 1300-1302. (10) Vettiger, P.; Cross, G.; Despont, M.; Drechsler, U.; Duerig, U.; Gotsmann, B.; Haberle, W.; Lantz, M.; Rothuizen, H.; Stutz, R.; Binnig, G. IEEE Trans. Nanotechnol. 2002, 1, 39-64. (11) Sheehan, P. E.; Whitman, L. J.; King, W. P.; Nelson, B. A. Appl. Phys. Lett. 2004, 85, 1589-1591. (12) Nelson, B. A.; King, W. P.; Laracuente, A.; Sheehan, P. E.; Whitman, L. J. Appl. Phys. Lett. 2006, 13, 033104. (13) Sunden, E. O.; Wright, T. L.; Lee, J.; Graham, S. A.; King, W. P. Appl. Phys. Lett. 2006, 88, 033107. (14) King, W. P.; Kenny, T. W.; Goodson, K. E. Appl. Phys. Lett. 2004, 85, 2086-2088. (15) Gotsmann, B.; Durig, U. Langmuir 2004, 20, 1495-1500. (16) Gotsmann, B.; Durig, U. Appl. Phys. Lett. 2005, 87. (17) Hammiche, A.; Reading, M.; Pollock, H. M.; Song, M.; Hourston, D. J. ReV. Sci. Instrum. 1996, 67, 4268-4274. (18) Zepeda-Ruiz, L. A.; Maiti, A.; Gee, R.; Gilmer, G. H.; Weeks, B. L. J. Cryst. Growth 2006, 291, 461-467. (19) Hammiche, A.; Hourston, D. J.; Pollock, H. M.; Reading, M.; Song, M. J. Vac. Sci. Technol., B 1996, 14, 1486-1491. (20) Fryer, D. S.; Nealey, P. F.; dePablo, J. J. Macromolecules 2000, 33, 6439-6447. (21) Alcoutlabi, M.; McKenna, G. B. J. Phys.: Condens. Matter 2005, 17, R461-R524.
NL061196P
2149
