Академический Документы
Профессиональный Документы
Культура Документы
Letter
pubs.acs.org/JPCL
3 Brian M. Foley,† Caroline S. Gorham,† John C. Duda,† Ramez Cheaito,† Chester J. Szwejkowski,‡
‡ ,§,∥
4 Costel Constantin, Bryan Kaehr,* and Patrick E. Hopkins*,†
†
5 Department of Mechanical and Aerospace Engineering, University of Virginia, 122 Engineer's Way, Charlottesville, Virginia 22904,
6 United States
‡
7 Department of Physics and Astronomy, James Madison University, 901 Carrier Drive, Harrisonburg, Virginia 22807, United States
§
8 Advanced Materials Laboratory, Sandia National Laboratories, 1515 Eubank SE, Albuquerque, New Mexico 87123, United States
∥
9 Department of Chemical and Nuclear Engineering, University of New Mexico, 209 Farris Engineering, Albuquerque, New Mexico
10 87106, United States
28
29
30
E nergy processes and vibrations in biological macro-
molecules such as proteins ultimately dictate biological,
chemical, and physical functions in living materials. The vast
over a range of length scales of the protein (a few angstroms to
the size of the protein itself, a few nanometers)6,13,14 that can
be fundamentally different from those found in polymers or
50
51
52
31 functional duties of proteins − from structural building blocks amorphous solids.15,16 In these protein systems, vibrations of 53
32 to molecular recognition, catalysis, and energy transduction1−5 the fractal structure can add additional complexities to 54
33 − are ultimately governed by the energetic vibrations of the understanding vibrational transport. Therefore, proteins 55
34 molecules. Thus, the connection between protein structure and represent an ideal material system to study the underlying 56
35 the pathways of energy flow is a fundamental concern that mechanisms driving vibrational thermal transport in a fractal 57
36 dictates protein functionality and reactivity. These factors drive solid. 58
37 a protein’s applicability and functionality for both native and The nature of the thermal conductivity of proteins was 59
38 artificial applications.6 studied computationally by Yu and Leitner via molecular 60
39 In addition to their fundamental importance in bioprocesses, dynamics simulations.13,17,18 They determined that anharmonic 61
40 proteins present a virtually unexplored avenue to study the vibrational interactions among protein molecules can open up 62
41 fundamentals of thermal transport in naturally occurring an additional channel for thermal conduction in a protein, in 63
42 nanostructured, percolating networks. To date, thermal line with previous theories of thermal conduction in fractal 64
43 conduction in biological materials has often been explored in objects.19−22 In two separate works, they also evaluated the 65
44 the context of amorphous solids and evaluated through thermal conduction in the case of a both spectral13 and gray18 66
45 comparison with some variation of the so-called “minimum mean free paths of protein vibrations. Unfortunately, no 67
46 limit to thermal conductivity”7−12 at elevated temperatures
47 above ∼50−100 K as well as data from atomistic molecular Received: January 26, 2014
48 dynamics simulations. However, the ribbon-like motifs of Accepted: March 7, 2014
49 proteins lead to self-similar structures and fractal-like objects
© XXXX American Chemical Society A dx.doi.org/10.1021/jz500174x | J. Phys. Chem. Lett. XXXX, XXX, XXX−XXX
The Journal of Physical Chemistry Letters Letter
156 measurements sample ∼50−100 nm beneath the Al/protein 11.39 MHz. This is the average heating in the volume sampled 219
157 interface. Therefore, the TDTR data were analyzed with a two- by TDTR (∼50−100 nm beneath the Al/protein interface).33 220
158 layer model of an Al film (the thickness of which was This estimated temperature rise is <1 K for all temperatures 221
159 determined from picosecond acoustics) on a semi-infinite and therefore negligibly affects our reported values. 222
160 protein layer. In addition, because of the combination of the The protein thermal conductivity as a function of temper- 223
161 pump-and-probe spot sizes and the pump modulation ature is shown in Figure 3. The thermal conductivity increases 224 f3
162 frequency, the heat transfer is nearly entirely 1-D in the
163 cross-plane direction.35 We determined the thermal con-
164 ductivity of protein films by fitting the ratio of the in-phase
165 to out-of-phase lock-in output to a two-layer axially symmetric
166 thermal model that accounts for pulse accumulation during the
167 experiment.33−35 We fit t he thermal model to the data by
168 varying the thermal conductivity of the protein. We assume
169 literature values for the heat capacity of Al42 and the heat
170 capacity of the protein. 43,44 We deduce the thermal
171 conductivity of the Al film via the Wiedemann−Franz Law
172 and electrical resistivity measurements. Because of the low
173 thermal conductivity of the proteins, our measurements were
174 completely insensitive to the Al/protein thermal boundary
175 conductance. Therefore, in our analysis, the only fitting
176 parameter was the protein thermal conductivity. We took five
177 to seven TDTR scans at different locations on four different
178 samples of BSA (two on silicon and two on cover glass) and
179 two different samples of myoglobin (both on silicon). The
180 uncertainty in our measurements is determined by considering
181 the repeatability of the measurements at the different sample
182 locations, uncertainty in the local Al film thickness determined Figure 3. Thermal conductivity of BSA and myoglobin (this work),
183 by picosecond ultrasonics, uncertainty in the thermal SiO2,47 and polystyrene (PS)45 as a function of temperature. The
184 conductivity of the Al film, and uncertainty in the heat increase in thermal conductivity in these systems over this temperature
range can be ascribed to anharmonic vibrational interactions.
185 capacities of the proteins. In general, the uncertainty in the
186 measured thermal conductivity of the protein films determined
187 from our TDTR measurements is ∼15%. with temperature, although the values begin to level off upon 225
188 The room-temperature thermal conductivities of our BSA approaching room temperature. We do not heat the sample 226
189 and myoglobin films are 0.231 ± 0.031 and 0.190 ± 0.024 W substantially above room temperature to ensure that we are not 227
190 m−1 K−1, respectively. The thermal conductivities of the BSA denaturing the proteins. For comparison, we also show the 228
191 samples on silicon and glass substrates did not exhibit any thermal conductivity of polystyrene (PS),45 which has been 229
192 difference within the experimental uncertainty. Our BSA previously analyzed in terms of anharmonic coupling of 230
193 measurements agree with previous measurement by Park et vibrations.46 The PS data show very similar trends to our 231
194 al.,24 who determined the thermal conductivity of BSA as 0.265 protein data, which suggest that anharmonic coupling of 232
195 ± 0.08 W m−1 K−1 from various measurements of BSA in water vibrations in the protein structure is contributing to thermal 233
196 solutions. conductivity. As another comparison, we show the thermal 234
197 Unlike measurements of the thermal conductivity of proteins conductivity of SiO2.47 This increase in thermal conductivity 235
198 in solution, our solid protein films allow us to measure the has also been analyzed in terms of anharmonic coupling of 236
199 thermal conductivity of the protein structure at different localized vibrations.46 In other words, much like SiO2 and PS, 237
200 temperatures without having to account for phase changes in anharmonic vibrational coupling contributes to the thermal 238
201 the solution (e.g., freezing of water). We mount our protein conductivity of proteins. 239
202 samples in a liquid-nitrogen-cooled cryostat with optical access In addition to demonstrating thermal conductivity measure- 240
203 for our laser and measure the thermal conductivities of ments of solid proteins over a range of temperatures, one of our 241
204 myoglobin and BSA from 77 K to room temperature. We goals of this work is to lend insight into previous theories on 242
205 only tested the thermal conductivities of the samples on silicon fracton transport. The various fracton theories presented by 243
206 substrates to minimize steady-state heating from the absorbed Orbach and colleagues19−22 claim that an increase in thermal 244
207 laser power.33 For the samples on the silicon substrates, we conductivity over this temperature range is partially due to 245
208 estimate that the steady-state heating from the absorbed laser anharmonic fracton hopping.20 However, the various amor- 246
209 will increase the sample temperature by <1 K over all phous materials that have been analyzed in terms of this fracton 247
210 temperatures. However, for the samples on glass substrates theory are not necessarily fractal, and they can also be well- 248
211 that we tested only at room temperature, we reduced the described by other theories such as the minimum limit to 249
212 incident power but still approximate the steady state temper- thermal conductivity.7−11 In fact, Freeman and Anderson48 250
213 ature rise as ∼20 K. As previously noted, we did not observe discuss that it is not clear what excitations are responsible for 251
214 any difference in measured thermal conductivity so assert that thermal transport in amorphous structures and Cahill and 252
215 this steady-state laser heating difference in the samples on Pohl49 conclude that the fracton theory does not apply to 253
216 different substrates did not change the thermal state of the purely amorphous solids and polymers. Therefore, it is difficult 254
217 material. We expect additional heating in the protein samples to draw any conclusions about the validity of the fracton 255
218 by the absorbed energy from the modulated pump pulses at theories based on previous experiments. However, protein 256
257 molecules are well-known fractal structures, and our data show not change with temperature according to Yu and Leitner’s 278
258 qualitatively that anharmonic coupling of vibrations in this results and previous simulations by Shenogin et al.13,18,46 As 279
259 fractal network contribute to thermal conductivity. We cannot already established, the increase we observe in thermal 280
260 conclude whether the thermal vibrations in these solid protein conductivity is attributed to anharmonic interactions of the 281
261 samples are indeed fractons or simply localized vibrations; vibrations in the solid, as further supported by the agreement 282
262 however, the observation of anharmonic-like trends in the between in our experimental data and the anharmonic models 283
263 thermal conductivity of these fractal structure will lend insight by Yu and Leitner. (The magnitude of the residuals between 284
264 into future works. our data and the various models is plotted in the bottom panel 285
265 To further the knowledge of how vibrations are contributing of Figure 4.) Because of the uncertainty in our data and our 286
266 to heat transport in fractal proteins, we compare our measured high-temperature limitations to ensure our proteins are not 287
267 thermal conductivity of myoglobin to the molecular dynamics denatured, we can not conclusively determine whether our data 288
f4 268 simulations from Yu and Leitner.13,18 In Figure 4, we plot the shows better agreement with the gray model18 from Yu and 289
Leitner than the spectral model.13 However, this same 290
temperature trend is observed in systems in which anharmoni- 291
cally interacting vibrations are localized to a single mean free 292
path on the order of the intermolecular spacings, such as PS 293
and SiO2 (ref 46). More research must be conducted to 294
determine the length scale of the localization of vibrations in 295
fractal proteins and whether these vibrations are indeed a 296
fracton. 297
The results and discussion above lend further insight into 298
how anharmonic coupling of localized vibrations contribute to 299
thermal conductivity. According to existing theories of thermal 300
conductivity that consider anharmonicity as an independent 301
scattering mechanism as opposed to an independent con- 302
duction channel, anharmonic scattering of vibrations (pho- 303
nons) leads to a decrease in thermal conductivity.50 Contrary to 304
this prediction, in the limit of atomic disorder, it has been 305
shown computationally that three-body interactions increase 306
the thermal conductivity of the vibrational system.51,52 These 307
results have been explained in terms of the anharmonic decay 308
of localized modes in disordered systems, which would 309
otherwise not diffuse thermally.13,18,20,51,52 Furthermore, 310
Leitner et al.53 theoretically estimated the vibrational energy 311
transfer of normal modes, localized and extended, in a 1-D 312
disordered glass finding that the anharmonic contribution to 313
thermal conductivity is provided almost in full by spatially 314
overlapping localized modes. This result is applicable to the 315
protein structure, a randomly arranged heteropolymer struc- 316
tured by bonds of wide-ranging strength,53 where most normal 317
modes of protein and its secondary structure have shown to be 318
Figure 4. (top) Thermal conductivity of myoglobin (squares) as a spatially localized.54 319
function of temperature compared with the various models reported
by Yu and Leitner.13,18 The increase in the thermal conductivity with
In summary, we measured the thermal conductivity of solid, 320
temperature indicates that anharmonic interactions are contributing to water-insoluble protein films of BSA and myoglobin from 77 to 321
the thermal conductivity. Furthermore, on the basis of Yu and 296 K. The thermal conductivities of the proteins increase with 322
Leitner’s computational findings, the observed plateau in the data at increasing temperature. These increasing temperature trends 323
higher temperatures supports the fact that the mean free paths of the indicate that anharmonic coupling of vibrations in the protein is 324
vibrations in proteins are localized to a common length scale. contributing to thermal conductivity. This is consistent with 325
(bottom) Residuals between measured experimental data and the four vibrational transport in nonfractal amorphous solids, such as 326
models shown in the top plot (filled squares: harmonic and constant SiO2 and PS. We cannot conclude whether the thermal 327
MFP; filled circles: harmonic and energy-dependent MFP; open vibrations in these solid protein samples are indeed fractons or 328
squares: anharmonic and constant MFP; open circles: anharmonic and
simply localized vibrations; however, the observation of 329
energy-dependent MFP).
anharmonic-like trends in the thermal conductivity of a fractal 330
structure will lend insight into future works and could form the 331
269 myoglobin data from Figure 3 along with the results from four basis for future validation of seminal theories from Orbach and 332
270 different simulations from Yu and Leitner accounting for: (i) colleagues.19−22 Future computational research could elucidate 333
271 harmonic interactions only with a constant mean free path (i.e., the length scale of the anharmonic vibrational localization 334
272 gray model); (ii) harmonic interactions only with a spectrally length in fractal proteins and whether this vibration is indeed a 335
273 dependent mean free path; (iii) harmonic and anharmonic fracton. 336
274 interactions with a gray model; and (iv) harmonic and Finally, future studies on the heat transport characteristics in 337
275 anharmonic interactions with spectrally dependent mean free solid protein films offer the potential to experimentally realize 338
276 paths. Over the temperature range investigated in our study, if novel phenomena in thermal conduction. For example, as 339
277 anharmonicity was nonexistent, the thermal conductivity would pointed out by Yu and Leitner,17 heat diffusion in a protein is 340
341 subdiffusive. Therefore, conductance measurements of protein (12) Lubchenko, V.; Wolynes, P. G. The Origin of the Boson Peak 402
342 films of various thicknesses could give experimental insight into and Thermal Conductivity Plateau in Low-Temperature Glasses. Proc. 403
343 the mechanisms of anomalous heat diffusion.55 Additionally, Natl. Acad. Sci. 2003, 100, 1515−1518. 404
344 proteins are complex network structures. Liu et al.56 have (13) Yu, X.; Leitner, D. M. Heat Flow in Proteins: Computation of 405
Thermal Transport Coefficients. J. Chem. Phys. 2005, 122, 054902. 406
345 reported that thermal transport in complex network structures
(14) Pereyra, M.; Mendez, E. In Application of Thermodynamics to 407
346 can give rise to asymmetrical thermal conductivity and non- Biological and Materials Science; Tadashi, M., Ed.; InTech: Rijeka, 408
347 Gaussian heat current distributions. Therefore, solid protein Croatia, 2011; Chapter 9, pp 243−258. 409
348 films could provide an avenue to realize these nontraditional (15) Losego, M. D.; Moh, L.; Arpin, K. A.; Cahill, D. G.; Braun, P. V. 410
349 regimes of thermal transport. Interfacial Thermal Conductance in Spun-Cast Polymer Films and 411
350
351
■ AUTHOR INFORMATION
Corresponding Authors
Polymer Brushes. Appl. Phys. Lett. 2010, 97, 011908.
(16) Cahill, D. G.; Lee, S. M.; Selinder, T. I. Thermal Conductivity of
κ-Al2O3 and α-Al2O3 Wear-Resistant Coatings. J. Appl. Phys. 1998, 83,
5783−5786.
412
413
414
415
352 *E-mail: bjkaehr@sandia.gov. (17) Yu, X.; Leitner, D. M. Anomalous Diffusion of Vibrational 416
353 *E-mail: phopkins@virginia.edu Energy in Proteins. J. Chem. Phys. 2003, 119, 12673−12679. 417
354 Notes (18) Yu, X.; Leitner, D. M. Vibrational Energy Transfer and Heat 418
Conduction in a Protein. J. Phys. Chem. B 2003, 107, 1698−1707. 419
355 The authors declare no competing financial interest.
■
(19) Orbach, R. Phonon Localization and Transport in Disordered 420
Systems. J. Non-Cryst. Solids 1993, 164−166 (Part 2), 917−922. 421
356 ACKNOWLEDGMENTS (20) Alexander, S.; Entin-Wohlma, O.; Orbach, R. Phonon-Fracton 422
357 P.E.H. appreciates financial support from the Office of Naval Anharmonic Interactions: The Thermal Conductivity of Amorphous 423
Materials. Phys. Rev. B 1986, 34, 2726−2734. 424
358 Research (N00014-13-4-0528). B.K. acknowledges support (21) Jagannathan, A.; Orbach, R.; Entin-Wohlman, O. Thermal 425
359 from the U.S. Department of Energy (DOE), Office of Science, Conductivity of Amorphous Materials above the Plateau. Phys. Rev. B 426
360 Office of Basic Energy Sciences (BES), Division of Materials 1989, 39, 13465−13477. 427
361 Sciences and Engineering. This work was partially supported by (22) Alexander, S.; Laermans, C.; Orbach, R.; Rosenberg, H. M. 428
362 the Commonwealth Research Commercialization Fund Fracton Interpretation of Vibrational Properties of Cross-Linked 429
363 (CRCF) of Virginia and the 4-VA mini-grant for university Polymers, Glasses, and Irradiated Quartz. Phys. Rev. B 1983, 28, 4615− 430
364 collaboration in the Commonwealth of Virginia. Sandia 4619. 431
365 National Laboratories is a multiprogram laboratory managed (23) Michnik, A. Thermal Stability of Bovine Serum Albumin DSC 432
366 and operated by Sandia Corporation, a wholly owned Study. J. Therm. Anal. Calorim. 2003, 71, 509−519. 433
367 subsidiary of Lockheed Martin Corporation, for the U.S. (24) Park, B. K.; Yi, N.; Park, J.; Choi, T. Y.; Lee, J. Y. Thermal 434
368 Department of Energy National Nuclear Security Adminis- Conductivity of Bovine Serum Albumin: A Tool to Probe the 435
Denaturation of Protein. Appl. Phys. Lett. 2011, 99, 163702. 436
369 tration under contract no. DE-AC04-94AL85000.
■
(25) Kong, J.-Y.; Miyawaki, O.; Nakamura, K.; Yano, T. The 437
“Intrinsic” Thermal Conductivity of some Wet Proteins in Relation to 438
370 REFERENCES their Average Hydorphobicity: Analysis on Gels of Egg-Albumin, 439
371 (1) Alberts, B. Essential Cell Biology, 3rd ed.; Garland Science: New Wheat Gluten and Milk Casein. Agric. Biol. Chem. 1982, 46, 789−794. 440
372 York, 2009; pp 119−170. (26) Lervik, A.; Bresme, F.; Kjelstrup, S.; Bedeaux, D.; Miguel Rubi, J. 441
373 (2) Kukura, P.; McCamant, D. W.; Yoon, S.; Wandschneider, D. B.; Heat Transfer in Protein-Water Interfaces. Phys. Chem. Chem. Phys. 442
374 Mathies, R. A. Structural Observation of the Primary Isomerization in 2010, 12, 1610−1617. 443
375 Vision with Femtosecond-Stiumulated Raman. Science 2005, 310, (27) Kawai, K.; Suzuki, T.; Oguni, M. Low-Temperature Glass 444
376 1006−1009. Transition of Quenched and Annealed Bovine Serum Albumin 445
377 (3) Engel, G. S.; Calhoun, T. R.; Read, E. L.; Ahn, T. K.; Mancal, T.; Aqueous Solutions. Biophys. J. 2006, 90, 3732−3738. 446
378 Cheng, Y. C.; Blankenship, R. E.; Fleming, G. R. Evidence for (28) Cooper, A. Protein Heat Capacity: An Anomaly That Maybe 447
379 Wavelike Energy Transfer Through Quantum Coherence in Photo- Never Was. J. Phys. Chem. Lett. 2010, 1, 3298−3304. 448
380 synthetic Systems. Nature 2007, 446, 782−786. (29) Yamazoe, H.; Tanabe, T. Preparation of Water-Insoluble 449
381 (4) Champion, P. M. Following the Flow of Energy in Biomolecules. Albumin Film Possessing Nonadherent Surface for Cells and Ligand 450
382 Science 2005, 310, 980−982. Binding Ability. J. Biomed. Mater. Res., Part A 2007, 86A, 228−234. 451
383 (5) Miller, R. J. D. Vibrational Energy Relaxation and Structural (30) Sreerama, N.; Venyaminov, S. Y.; Woody, R. W. Estimation of 452
384 Dynamics of Heme Proteins. Annu. Rev. Phys. Chem. 1991, 42, 581− Protein Secondary Structure from Circular Dichroism Spectra: 453
385 614. Inclusion of Denatured Proteins with Native Proteins in the Analysis. 454
386 (6) Leitner, D. M. Energy Flow in Proteins. Annu. Rev. Phys. Chem. Anal. Biochem. 2000, 287, 243−251. 455
387 2008, 59 (31) Takeda, K.; Wada, A.; Yamamoto, K.; Moriyama, Y.; Aoki, K. 456
388 (7) Cahill, D. G.; Pohl, R. O. Lattice Vibrations and Heat Transport Conformational Change in Bovine Serum Albumin by Heat 457
389 in Crystals and Glasses. Annu. Rev. Phys. Chem. 1988, 39, 93−121. Treatment. J. Protein Chem. 1989, 8, 653−659. 458
390 (8) Cahill, D. G.; Watson, S. K.; Pohl, R. O. Lower Limit to the (32) Cahill, D. G.; Watanabe, F. Thermal Conductivity of 459
391 Thermal Conductivity of Disordered Crystals. Phys. Rev. B 1992, 46, Isotopically Pure and Ge-Doped Si Epitaxial Layers from 300 to 550 460
392 6131−6140. K. Phys. Rev. B 2004, 70, 235322. 461
393 (9) Cahill, D. G.; Pohl, R. O. Heat Flow and Lattice Vibrations in (33) Cahill, D. G. Analysis of Heat Flow in Layered Structures for 462
394 Glasses. Solid State Commun. 1989, 70, 927−930. Time-Domain Thermoreflectance. Rev. Sci. Instrum. 2004, 75, 5119− 463
395 (10) Hopkins, P. E.; Piekos, E. S. Lower Limit to Phonon Thermal 5122. 464
396 Conductivity of Disordered, Layered Crystals. Appl. Phys. Lett. 2009, (34) Schmidt, A. J.; Chen, X.; Chen, G. Pulse Accumulation, Radial 465
397 94, 181901. Heat Conduction, and Anisotropic Thermal Conductivity in Pump- 466
398 (11) Hopkins, P. E.; Beechem, T. Phonon Scattering and Velocity Probe Transient Thermoreflectance. Rev. Sci. Instrum. 2008, 79, 467
399 Considerations in the Minimum Phonon Thermal Conductivity of 114902. 468
400 Layered Solids Above the Plateau. Nanoscale Microscale Thermophys. (35) Hopkins, P. E.; Serrano, J. R.; Phinney, L. M.; Kearney, S. P.; 469
401 Eng. 2010, 14, 51−61. Grasser, T. W.; Harris, C. T. Criteria for Cross-Plane Dominated 470