key: cord-0721404-r3aevb93 authors: Qiu, Lin; Ouyang, Yuxin; Feng, Yanhui; Zhang, Xinxin; Wang, Xiaotian title: In vivo skin thermophysical property testing technology using flexible thermosensor-based 3ω method date: 2020-10-09 journal: Int J Heat Mass Transf DOI: 10.1016/j.ijheatmasstransfer.2020.120550 sha: 14b2dd7ffaf4f1dd756c9151e8e097df5ea9a216 doc_id: 721404 cord_uid: r3aevb93 Thermophysical properties of human skin surface and subsurface can reflect the hydration state of the skin and the blood flow rate in the near surface microvessels, which reveals important physiological information related to dermatology and overall health status of human body. Although a few techniques have been developed to measure these signs, complicated devices are required and the subjects need to be completely fixed during the test period. Here, a flexible thermosensor-based 3ω technology was used to monitor thermal conductivity of human skins at different states. Through the analysis of these characteristics, the corresponding physiological state can be established, which can provide a new detection method for the evaluation or prediction of human health status. As an important organ of the human body, the skin is a physical barrier that interacts with the surrounding environment to protect the internal organs of the body and is the first line of defense against external influences [1, 2] . The skin has rich perceptual ability, and can regulate the body temperature through the perfusion rate regulation of blood flow in the epidermal structure according to the balance between the body and the external environment [3] [4] [5] [6] . At the same time, the blood flow rate and hydration status under the skin can in turn display the health-related status information [7] [8] [9] [10] . Therefore, more and more health monitoring devices that detect the state of the physical condition through the skin are being developed. Skin lesions can reflect physiological information of the human body, thereby helping to quickly diagnose the disease. For traditional Chinese medical science, observing the skin condition is one of the main diagnostic methods, which was widely used in ancient China. In modern medical diagnosis, the correlation between skin lesions and diseases is gradually being explored by researchers. For example, when the influenza like illness broke out in British Columbia, Canada in March 2014, the rash was considered to be an immunological mechanism associated with influenza [11] . In the first half of 2020, New Coronary Pneumonia (COVID-19) is rampant in Asia, Europe and the Americas. There is also a large-scale crisis in Africa. This infectious disease is usually accompanied by fever, cough, chest tightness and other symptoms, and even produces asymptomatic changes, which makes the testing technology facing challenges [12] [13] [14] . As the first line of defense against these "global wars", reliable testing technology is in urgent need of continuous improvement and updating. In traditional medical field, the measurement of skin temperature is usually used to obtain general information about whether the human body maintains normal physiological characteristics as a whole, and the thermal conductivity (TC) of the skin can further express its internal tissue activity, degree of hydration and other information, which can further provide more information for medical diagnosis. Electronic skin is a product that develops based on skin bionics. It is an artificial, intelligent electronic device that mimics real skin function [15] . It senses physiological information through sensors and converts it into electronic signals for transmission and analysis. It is expected to achieve monitoring and diagnosis of human health. The most basic characteristic for an electronic skin is flexible and stretchable [16] . This allows the device to be placed against non-planar skin for more accurate measurements. At present, the commonly used method is to attach the low modulus lamellar array combination sensor to the skin, then they measure and monitor the tissue TC, subcutaneous blood flow changes and muscle activity through thermoelectric induction and biochemical sensing [8, 9, 17] . These non-invasive characterizations with simple devices requirement has changed the condition of the subject's limited action by the measuring instrument, enabling continuous measurement of the human body's movement.The 3 ω method has frequency domain characteristics, so the thermophysical properties of tested objects can be obtained by analyzing the relationship between the the temperature change response and the thermal wave frequency. The thermosensor is used as a heater emitting thermal wave and also a temperature sensor [18] [19] [20] [21] [22] . The unique advantage of this method is that it can control the thermal wave penetration depth by changing the alternating current (AC) frequency, so as to realize the layered measurement of multi-layer materials [23] . The development of flexible freestanding thermosensor remarkably improves the 3 ω method, which can be reused and achieve nondestructive measurement [24] [25] [26] . Therefore, it is reliable and creative to apply this method to the measurement of human skin with multi-layer structure in vitro and extract the thermophysical property information of each tissue layer. In this study, we achieved a non-destructive measurement of the thermal properties of human skin by using a 3 ω method based on the freestanding thermosensor. Through the measurement of different body parts and different states, the relationship between skin blood flow and TC is proved, which provides a new scheme for human physiological health detection. The traditional 3 ω method requires the deposition of miniature metal thermosensors with a specific size and shape on the surface of the samples to obtain thermal properties [27] . The measurement of the skin of the human body obviously makes it impossible to prepare the thermosensor using the deposition method. Therefore, in this study, the measurement was done by using the freestanding thermosensor made of a flexible film-wrapped metal strip. In addition to the non-destructive measurement of the human body, this method can also solve the problems of repeated preparation of micro-thermosensors, difficulty in insulation, and high surface roughness of the sample in the measurement of other materials [28] . In this study, measurements can be made by simply attaching the thermosensor to the specific part to be measured using a piece of medical tape. This is also an important advancement in the introduction of 3 ω thermophysical property measurement methods into human health monitoring. For the traditional measurement, it needs to deposit the thermosensor on the sample. An AC with a direct current (DC) offset component at an angular frequency of 1 ω is introduced to the deposited miniature metal thermosensor. Since the metal thermosensor itself has an electrical resistance, after the current is supplied, the heat generated by the Joule effect will heat the material at a frequency of 2 ω, and the material absorbs heat to generate a temperature wave at the same angular frequency. When the thermosensor emits the thermal wave, its own temperature rises and the resistance increases correspondingly. The change in the frequency of the increase in resistance is combined with an AC of frequency 1 ω to produce a harmonic voltage of 3 ω, and this third harmonic component is only related to the thermal action [29] . When the metal strip is deposited in the packaging film, it becomes the freestanding thermosensor. The freestanding thermosensor is fabricated using photolithography and UV exposure processes based on a flexible substrate film. The Kapton (Ka) film produced by DuPont TM was used as the flexible encapsulation film of the thermosensor. This kind of film has a very thin thickness of about 55 μ m (the measurement uncertainty is 2.0%), so that the thermal wave emitted by the metal thermosensor can easily penetrate the package film into the sample to be tested. Here nickel is used to deposit the miniature metal strip structure with four pads on a flexible substrate because it has a higher temperature coefficient of resistance (0.0064 K −1 , higher than platinum resistance and the measurement uncertainty is 2.5%) [30] which makes the thermosensor more sensitive. The metal strip used in this experiment has an in-between length of 7 mm (the measurement uncertainty is 1.0%) and the metal deposit layer has a thickness of 100 nm (the measurement uncertainty is 2.0%). The width is 600 μ m (the measurement uncertainty is 2.6%), and the value is in the range of optimum widths at which various materials are measured with high precision optical microscope [28] . After the metal strip is deposited, metal wires are taken out from the four pads for the connection to the instrument during measurement. Finally, the above mentioned structure is packaged by a hot pressing process using a Ka film with hot melt adhesive. The flexible freestanding thermosensor has high sensitivity and can be used repeatedly for thermophysical property measurement. Human experiments are carried out with the knowledge of the person (age: 24, female) being tested for the research of wearable sensor technologies. Moreover, the experiment has no damage to the human body except for a slight fever in the tested site. All participants was provided written informed consent before enrolment. During the measurement, the thermosensor attached to the skin is connected to the test system as shown in Fig. 1 . The input signal under different frequency is controlled by adjustable resistance and lock-in amplifier, then the harmonic signal is extracted and amplified by the amplifiers including signal generator, differential amplifiers and lock-in amplifier, and the data such as the TC is obtained by analyzing the experimental data. Based on the uncertainty of size measurement, and the uncertainty from these devices which are the range of 0.07%-0.16%, the comprehensive uncertainty of thermosensor measurement is 9.1%. Since the thermosensor is packaged using two flexible films, it is necessary to measure at low frequencies to ensure that the thermal wave completely penetrates the thermosensor structure into the body tissue to be tested. The thermal wave penetration depth is defined as: In this study, α is the thermal diffusivity of skin, κ is the TC and ρ m is the density of human skin. Table 1 gives the reference values of human skin that can be used to estimate the depth of thermal penetration for independent skin detection. When the measurement frequency is less than 1 Hz, the thermal penetration depth will exceed 70 μ m. This indicates that at lower frequencies the thermal wave is sufficient to penetrate the packaging film. The Table 1 Different thermophysical parameters of skin tissue structure [31, 32] . thickness of the package film is approximately 55 μ m, so when the thermosensor is placed on human skin for measurement, it can be regarded as a two-layer system on a semi-infinite surface. Simultaneously, the width of the nickel metal strip is much larger than the thickness of the film, and the heat flow can be considered as one-dimensional. Therefore, according to the semi-infinite and the multi-layer structure conduction formula derived by Borca et al. [33] , the temperature rise formula in the independent sensor structure can be described as: in where The subscript x, y denote the in-plane and out-of-plane direction, respectively, f denotes the flexible film, s denotes the sample to be tested, which is the skin in this study, k is the integral factor, b is the half width of the nickel band. p is power, and d is thickness. When using the flexible probe to measure samples, the thermal contact resistance between the thermosensor and the sample is a factor to be considered. When the temperature rise T obtained by the experimental measurement is separated into two parts from the sample (including the Ka film wrapped with the metal strip) and the thermal contact resistance R contact , the Eq. (2) can be corrected by referring to the Eq. (8) . where α CR is the temperature coefficient of resistance of nickel. U 3 ω and U 1 ω are the third harmonic voltage and fundamental voltage amplitude respectively. According to the measured temperature rise result, the theoretical temperature rise is taken as the fitting objective function, and the TC of the sample to be tested can be extracted by the least squares fitting of experimental temperature rise. It is the core for 3 ω technology to use temperature rise on thermosensor to gain thermophysical property of sample, and the temperature rise come from the Joule heat induced by fundamental voltage input. Based on Finite Element Method (FEM) analysis software COMSOL Multiphysics ® [34] , a model which exactly simulates the size of the test structure and the physical properties of the sample was built. The heat transfer process in the frequency domain is introduced to analyze the temperature rise of the thermosensor. In the light of Figure 2 a, the thickness of the tissues structure consisting of epidermis (Ep), papillary dermis (Pa), reticular dermis (Re), subcutaneous fat (Su) and Muscle (Mu) were set to be normal human level ( Table 1 ) as the initial value, which will change as the simulation progresses to match the real situation of the subject. The strip section between two inner electrodes is the heat source defined by Q = p/V, where Q is heat dissipation in the strip, p and V are inputting power and volume of the strip, respectively. Harmonic perturbation model which predicts heat transfer process in frequency domain is an important approach for the temperature rise analysis. Specifically, when subjected to periodic sinusoidal heating under a given frequency, the temperature response of the micro thermosensor contacting with the skin is also a periodic, sinusoidal and with double frequency around an equilibrium value. The surface of thermosensor and the skin were initially given heat transfer conditions identical to actual situation, including low natural convection effect of air and radiation effects in the laboratory. Considering the topical influence from the thermosensor, the flanks of the skin were set to be adiabatic condition, and the bottom of the skin maintained constant temperature which was close to normal human body temperature. In summary, by modulating the thermal wave penetration depth to the skin using the changeable AC frequencies, the temperature rise of the thermosensor give a direct feedback of the thermophysical properties of skin inner layers. Since the fitting model includes skin TC, specific heat and multiple unknown physical parameters related to the encapsulation film, in order to improve the fitting accuracy, the sensitivity analysis S x of the model at different frequencies is first performed. it expresses the sensitivity of the temperature rise T under the independent variable x to determine the performance of the thermosensor. The fixing tools such as tape are used to fix the thermosensor on the surface of the standard sample. It should be noted that the tape does not totally touch the part where the thermosensor generates heat and receives heat signals when the thermosensor is fixed to the skin. By controlling the input voltage, the thermosensor emits thermal wave which penetrates into the sample and the temperature rise of the thermosensor is a feedback of the sample's TC. The temperature rise model is output to the display device in the form of electrical signal (third harmonic voltage and phase information). The sensitivity of the thermal resistance between the skin and the thermosensor, the TC of the skin, and the TC of the thermosensor encapsulation film to the temperature rise at different frequencies were calculated based on Eq. (10) . As summarized in Fig. 2 b, the sensitivity of the thermal resistance in the frequency range above 10 Hz is significantly higher than the other two parameters, and the sensitivity of the TC of the encapsulation film below 1 Hz is higher than the high frequency range. The sensitivity of TC of the sample to be tested is not high overall, which is only sensitive below 1 \ ,Hz. Therefore, the data fitting extraction of the TC of the sample to be tested is performed at last. The reliability of the measurement method is verified by measuring standard sample bands with known TC such as quartz glass, stainless steel and adhesive before skin measurement. Fig. 2 c shows that the measurement signals of different samples differ significantly, which is related to the characteristics of the samples themselves. The extracted TC using the 3 ω method based on the freestanding sensor was compared with the reference value ( Fig. 2 d) [31, 35, 36] , and the correlation coefficient was 0.99, which indicated that the 3 ω method and the mathematical model used in this research were credible. We measured several parts of the body skin which have significantly different subcutaneous blood flow and fat thickness, including the forehead, the back of the hand, the cheek, and the inside of wrist ( Fig. 2 e) . The curvature of these parts is relatively flat and the core measurement part of thermosensor is small (length about 7 mm), that is why there are negligible impact on measurement from slight bending. Apart from that, the measurement of the wrist was tested under normal state, with rash state and scratching after the rash ( Fig. 2 f) . The person being measured is a 24-year-old female with a body mass index of 17.9 kg/m 2 . It should be noted that the rash is caused by seasonal allergic dermatitis, and the scratching strength of the skin is not strong. Hence, there is no obvious skin redness. The results show that the TC of the cheek skin was 0.49 W/(m · K), which was higher than the skin area of the forehead (0.34 W/(m · K)) ( Fig. 3 ). Fig. 3 shows the TC values measured by the flexible sensor at different skin areas. It should be noted that the thermal wave penetration depth generated by the thermosensor in the skin cannot be shown based on the results alone, because these TC values are comprehensive values of multiple tissues including skin, subcutaneous fat, and muscle. However, it can be seen from the results that the difference in organism plays a significant role. The cheek has a thicker layer of tissue and a higher hydration, so its TC is greater than that of the forehead. The results of three different states on the inner side of the wrist are shown in Fig. 3 . It can be seen that the TC of the same area of the skin is not much different, which is in line with com-mon sense. In areas where the rash has grown or been scratched, there is an increase in TC. Due to the problem of measuring the depth, it is difficult to obtain the effect of the rash only by this result, because this change may also include the effect of Su. The subsequent studies will further analyze these results. In order to refine the connection between the measurement data and the skin and related tissues, the simulation research based on the FEM feedbacks the experimental measurement situation and further extracts the effective information of the experimental data. The reliability of simulation model is verifed by comparing with experimental results, the measurement data of the back of the subjects hand is selected for the analysis. Fig. 4 a illustrates the temperature of the thermosensor and the skin model from the FEM simulation. It can be clearly seen that the temperature of the Ep layer is lower than the normal human temperature due to the influence of the lower room temperature (about 293 K). In order to better distinguish the temperature difference between the thermosensor and the skin, we further extract the surface position and expand the temperature range. It can be considered that under a good fit, the temperature of the non-heat source part of the thermosensor and the skin temperature are basically the same (within 0.27 K), which greatly reduces the interference of heat transfer from convection and thermal radiation. According to the working principle of the thermosensor, the metal strip of the thermosensor generates heat which diffusese into the skin. Under the premise of controlling the input power (input voltage is only 180-200 mV), the temperature rise of the metal strip induced by Joule effect is not large, so as to realize the non-destructive measurement of the skin. Therefore, the temperature of the heat source part of the thermosensor is higher than the skin temperature. The infrared image collected by VarioCAM®HD research 600 demonstrates that the temperature distribution of the skin surface is generally uniform after a steady state was reached ( Fig. 4 b) . Since the temperature bar range measured by thermal imaging is relatively large, there is little difference between the two in the obtained thermal imaging picture. The temperature of the metal strip is slightly higher than the skin from extraction of temperature in the image, which well verifies the rationality of the FEM simulation. The temperature oscillations of the thermosensor under different frequencies are presented in Fig. 4 c, it means that the heat is transmitted in a weak but stable form from the strip into the skin when the system reach steady state. The peaks represent the maximum distance of thermal penetration into the skin, that is the temperature rise T . It is clearly shown that the temperature rise degree of the strip decreases with the decreasing frequencies, which is similar to experiment results. T values of the simulation generally agree well with the experimental data ( Fig. 4 d) . The slight difference is due to the unstable heat transfer with the environment during the experimental process, which comes from realtime and slight changes in the convective heat transfer coefficient and temperature in the experimental environment. However, this change has little effect on the thermal wave penetration depth in the skin obtained by simulation prediction and experimental measurement [25] . To vividly reflect experimental measurements, thermal penetration depths under different frequencies are presented in Figure 4 e. It is clearly seen that the maximum transmission distance of thermal wave from the thermosensor to the skin is close to 8 mm, which can only be achieved in experimental measurements at most ~6 mm as the temperature rise greatly decreases with depth [25] . This is also shown in the result from Fig. 4 e. The thermal wave originated from the thermosensor sequentially penetrates Ka layer, Ep layer, Pa layer, Re layer and Su layer under low AC frequencies. When the frequency exceeds 90 Hz, the heat is enclosed in the thermosensor, and begins to penetrate to the skin as the frequency falls below 90 Hz. When the frequencies reach 10 Hz, 0.6 Hz, 0.2 Hz, and 0.01 Hz, respectively, the heat has the intensity to penetrate the corresponding skin Pa, Re, Su and Mu layers, respectively ( Fig. 5 ) . Based on the thermal penetration depth versus frequency curve, the measurement data are split into four frequency bands to extract the TC of specific layers, i.e ., 10-10 0 0 Hz, 0.6-10 0 0 Hz, 0.2-10 0 0 Hz and 0.01-10 0 0 Hz, as illustrated in Fig. 5 b, d, f and h. It is interesting to note that the thermal penetration designated layers can be easily obtained by performing least-square-fitting using different frequency bands as mentioned above. The poor fitting in 10-10 0 0 Hz frequency band is due to the weak detective sensitivity of the micro thermosensor for the ultra-thin object, i.e ., ~75 μ m thick Ep layer, which is close to that of the micro thermosensor ( ~100 μ m). To extract the TC value of each layer, the following expression for the thermal resistance in series is used: where d denotes thickness ( μ m), κ denotes TC (W/(m · K)), R denotes thermal resistance (m 2 · K/W). Subscript object represents the overall from Ka layer to the entire skin. R contact represents the thermal contact resistance between the micro thermosensor and the skin. According to the experimental measurement and preliminary research, it is known that the R contact is 0.0 014 ± 0.0 0 01 m 2 · K / W , d Ka and κ Ka is 55 ± 5 μ m and 0.83 ± 0.03 W / (m · K) , respectively [20, 21, 24] . Combined with real thickness data of each layer of the skin, the TC of each skin layer are 0.52 ± 0.03 W / (m · K) , 0.84 ± 0.08 W / (m · K) , 0.90 ± 0.08 W / (m · K) and 0.42 ± 0.03 W / (m · K) ( Table 2 ). It can be seen that the TC of the Pa layer and the Re layer are not different, the reason is that the two belong to the dermis layer and have similar properties. Due to the different physiological characteristics of different human bodies, the skin TC will change to some extent, which makes the calculation results different from some research results, but the proportion of TC in the skin is basically similar [31, 32] . According to this process, the forehead, cheeks, and wrists (corresponding to three different situations) are analyzed and simulated for thermal penetration depth, then the corresponding experimental data is piecewise fitted to obtain TC values at different organization levels in these parts. The results are also shown in Table 2 . Under the segmented analysis of the experimental data, each layer of the skin has certain commonality in the measurement frequency band. The Ep corresponds to the frequency range from two digits to three digits, the dermis layer (Pa + Re) is a frequency segment above 0.1 Hz, and it is difficult to maintain the maximum penetration distance of the heat wave in the Su layer below 0.01 Hz. When the frequency parameter is set in 10 −2 magnitude and even lower, the heat wave has the ability to reach the muscle layer. This provides a measurement benchmark for our skin measurement systems. In addition, different environmental conditions (convection heat exchange intensity, radiation heat exchange intensity, and ambient temperature) during the simulation process have a weak influence on the intensity of the heat wave generated by the thermosensor, but the contact effect between the thermosensor and the skin can greatly affect the heat wave intensity, which provide a direction for study of the next complete thermosensor. Combining the results of most researchers, the effective TC of the skin is in the range 0.28 ± 0.03-0.73 ± 0.14 W / (m · K) when subcutaneous fat is considered [37] . The results of this study are basically within this range, the numerical difference comes from the difference in skin structure and condition. This includes the thickness of the internal tissue of the skin, the moisture content, temperature, and blood perfusion rate of the capillaries produced by the physiological conditions of the human body. According to the research by Okabe et al. [38] , skin thickness and moisture content have a greater influence on the test results. Although there is a certain threshold effect for the increase in skin thickness, the change in moisture content makes the epidermal TC constantly changing, which leads to different measurement results in the different skin parts and in the same part at different times. Rash is a common skin lesions, which is usually generated in the Ep and dermis, and difficult to reach the Su layer. According to the conclusions of some studies [9, 39] , Fig. 6 shows the physiological changes of the specific tissues of the skin under these three conditions. After peeling off the TC of the Su layer at the wrist, the TC at the normal conditions, the presence of the rash and the scratching of the rash were 0.41 ± 0.03 W / (m · K) , 0.40 ± 0.03 W / (m · K) and 0.45 ± 0.04 W / (m · K) , respectively. Although the value did not change much, which is in line with common sense in test of same measurement area on the wrist, the TC changed a lot in Ep because of the characteristic of the rash (The specific tissues of the TC change results are shown in Table 2 ), that is, there are multiple apophyses on Ep, so that there are a lot of gaps between the apophyses, which is equivalent to porous materials. This structural feature weakens the heat transfer ability of Ep, and the TC decreased from 0.32 ± 0.02 W / (m · K) under normal conditions to 0.14 ± 0.01 W / (m · K) with rash symptoms. However, after the scratching, the subcutaneous blood flow velocity increased and the TC of Ep layer increased from 0.14 ± 0.01 W / (m · K) to 0.17 ± 0.01 W / (m · K) , and the TC of the dermis layer remained basically unchanged [9] , so theoverall effective value increased (from 0.40 ± 0.03 W / (m · K) to 0.45 ± 0.04 W / (m · K) ). Although the rash is not generated by Su, due to many micronutrient deficiencies that may exist in this lesion, there is a loss of Su in local locations [39] , which caused the TC of this layer to change differently than normal condition, the TC changed from 0.36 ± 0.03 W / (m · K) (normal condition) to 0.86 ± 0.08 W / (m · K) (rash symptoms) and 0.65 ± 0.06 W / (m · K) (rash symptoms and scratching). It can be speculated that scratching reduces the loss effect of Su layer. This paper achieved the non-destructive measurement of the 3 ω method on the living body. Using the freestanding thermosensor, the monitoring of human physiological conditions can be conveniently realized. Combined with the finite element analysis, the TC value of each layer including the skin is extracted, which makes the analysis of the skin thermal characteristics more detailed. It can be known from the measurement that the TC of the wrist reaches 0.45 W / (m · K) after scratching due to the increased blood flow, which is higher than that of the wrist under normal conditions. The 3 ω method can detect this change sensitively, which provides a new solution for future health detection technologies. The authors declare that we have no financial and personal relationships with other people or organizations that can inappropriately influence our work, there is no professional or other personal interest of any nature or kind in any product, service and/or company. The skin: an indispensable barrier Barrier function of the skin:la raison d'etre of the epidermi Neuronal control of skin function: the skin as a neuroimmunoendocrine organ Skin temperature: its role in thermoregulation Control of skin blood flow in the neutral zone of human body temperature regulation Epidermis of human skin: pyroelectric and piezoelectric sensor layer X-ray videodensitometric methods for blood flow and velocity measurement: a critical review of literature Ultrathin conformal devices for precise and continuous thermal characterization of human skin Epidermal devices for noninvasive, precise, and continuous mapping of macrovascular and microvascular blood flow Resting blood flow in the skin: does it exist, and what is the influence of temperature, aging, and diabetes? Case series of rash associated with influenza b in school children Who launches crowdfund for COVID-19 response As normalcy returns, can china keep COVID-19 at bay? Countries test tactics in war against COVID-19 25th anniversary article: the evolution of electronic skin (e-skin): a brief history, design considerations, and recent progress Flexible electronics: the next ubiquitous platform Thermal conductivity of amorphous solids above the plateau 3 ω method for specific heat and thermal conductivity measurements Remarkably enhanced thermal transport based on a flexible horizontally-aligned carbon nanotube array film Iodine nanoparticle-enhancing electrical and thermal transport for carbon nanotube fibers Ultrasensitive and stretchable temperature sensor based on thermally stable and self-healing organohydrogels Application of the three omega thermal conductivity measurement method to a film on a substrate of finite thickness Thermal transport barrier in carbon nanotube array nano-thermal interface materials Inhomogeneity in pore size appreciably lowering thermal conductivity for porous thermal insulators Extremely low thermal conductivity of graphene nanoplatelets using nanoparticle decoration Thermal conductivity measurement from 30 to 750 k: the 3 ω method Design and application of a freestanding sensor based on 3 ω technique for thermal-conductivity measurement of solids, liquids, and nanopowders Analysis of temperature fluctuations in ac heated filaments Temperature coefficients of resistance of metallic films in the temperature range 25 to 600 • C Quantification of the thermal signature of a melanoma lesion A finite element model of skin subjected to a flash fire Data reduction in 3 ω method for thin-film thermal conductivity determination CRC handbook of chemistry and physics A practical extension of the 3 ω method to multilayer structures In vivo thermal conductivity of the human forearm tissues Non-invasive measurement of effective thermal conductivity of human skin with a guard-heated thermistor probe Nutrition in the Prevention and Treatment of Disease This work received financial support from Beijing Nova Program ( Z20110 0 0 06820 065 ), National Natural Science Foundation of China ( 51876008 , 51876 007 ), Beijing Natural Science Foundation ( 3202020 ), Interdisciplinary Research Project for Young Teachers of USTB (Fundamental Research Funds for the Central Universities ) ( FRF-IDRY-19-004 ).