Search

High-volume manufacturing of devices based on transition metal dichalcogenide (TMD) ultrathin films will require deposition techniques that are capable of reproducible wafer-scale growth with monolayer control. To date, TMD growth efforts have largely relied upon sublimation and transport of solid precursors with minimal control over vapor-phase flux and gas-phase chemistry, which are critical for scaling up laboratory processes to manufacturing settings. To address these issues, we report a new pulsed metal–organic chemical vapor deposition (MOCVD) route for MoS2 film growth in a research-grade single-wafer reactor. Using bis(tert-butylimido)bis(dimethylamido)molybdenum and diethyl disulfide, we deposit MoS2 films from ∼1 nm to ∼25 nm in thickness on SiO2/Si substrates. We show that layered 2H-MoS2 can be produced at comparatively low reaction temperatures of 591 °C at short deposition times, approximately 90 s for few-layer films. In addition to the growth studies performed on SiO2/Si, films with wafer-...

A nondispersive infrared (NDIR) gas analyzer was demonstrated for measuring the vapor-phase density of the carbonyl-containing organometallic cobalt precurso μ2-η2-(tBu-acetylene) dicobalthexacarbonyl (CCTBA). This sensor was based on direct absorption by CCTBA vapor in the C≡O stretching spectral region and utilized a stable, broadband IR filament source, an optical chopper to modulate the source, a bandpass filter for wavelength isolation, and an InSb detector. The optical system was calibrated by selecting a calibration factor to convert CCTBA absorbance to a partial pressure that, when used to calculate CCTBA flow rate and CCTBA mass removed from the ampoule, resulted in an optically determined mass that was nominally equal to a gravimetrically-determined mass. In situ Fourier transform infrared (FT-IR) spectroscopy was performed simultaneously with the NDIR gas analyzer measurements under selected conditions in order to characterize potential spectroscopic interferences. Interference due to CO evolution from CCTBA was found to be small under the flow conditions employed here. A CCTBA minimum detectable molecular density as low as ≈3 × 1013 cm−3 was calculated (with no signal averaging and for a sampling rate of 200 Hz). While this NDIR gas analyzer was specifically tested for CCTBA, it is suitable for characterizing the vapor delivery of a range of carbonyl-containing precursors.

A nondispersive infrared gas analyzer was demonstrated for investigating metal alkylamide precursor delivery for microelectronics vapor deposition processes. The nondispersive infrared analyzer was designed to simultaneously measure the partial pressure of pentakis(dimethylamido) tantalum, a metal precursor employed in high volume manufacturing vapor deposition processes to deposit tantalum nitride, and dimethylamine, the primary decomposition product of pentakis(dimethylamido) tantalum at typical delivery conditions for these applications. This sensor was based on direct absorption of pentakis(dimethylamido) tantalum and dimethylamine in the fingerprint spectral region. The nondispersive infrared analyzer optical response was calibrated by measuring absorbance as a function of dimethylamine and pentakis(dimethylamido) tantalum density. The difference between the mass of material removed from the ampoule during flow tests as measured gravimetrically and as determined optically, by calculating flow rates from the nondispersive infrared analyzer measurements, was only ≈2 %. The minimum detectable molecular densities for pentakis(dimethylamido) tantalum and dimethylamine were ≈2 × 1013 cm−3 and ≈5 × 1014 cm−3, respectively (with no signal averaging and for a sampling rate of 200 Hz), and the corresponding partial pressures were ≈0.1 Pa and ≈2 Pa for pentakis(dimethylamido) tantalum and dimethylamine, respectively (for an optical flow cell temperature of 93 ℃). Pentakis(dimethylamido) tantalum could be detected at all conditions of this investigation and likely the majority of conditions relevant to high volume manufacturing tantalum nitride deposition. Dimethylamine was not detected at all conditions in this study, because of a lower nondispersive infrared analyzer sensitivity to dimethylamine compared to pentakis(dimethylamido) tantalum and because conditions of this study were selected to minimize DMA production. While this nondispersive infrared gas analyzer was specifically developed for pentakis(dimethylamido) tantalum and dimethylamine, it is suitable for characterizing the vapor delivery of other metal alkylamide precursors and the corresponding amine decomposition products, although in the case of some metal alkylamides a different bandpass filter would be required.

Low volatility precursors are widely utilized in chemical vapor deposition (CVD) and atomic layer deposition (ALD) processes. Compared to gases and high volatility liquid precursors, delivery of low volatility liquid and solid precursors can be problematic, with solid precursors being particularly so. To investigate some of these delivery issues, the performance of a vapor draw vessel was characterized for the delivery of pentakis(dimethylamido) tantalum (PDMAT), a low-volatility solid precursor at preferable delivery temperatures, for reduced-pressure cyclical CVD and ALD processes. Vessel characterization involved determining (1) a source efficiency as a function of process conditions and (2) the degree of PDMAT decomposition as a function of temperature and vessel idle time. The PDMAT partial pressure, flow rate, and mass per injection used to determine the source efficiency were determined from measurements obtained using a custom-designed non-dispersive infrared gas analyzer. For a series of injections after an idle/purge sufficiently long to saturate the vessel head space, the source efficiency decreased from a maximum slightly less than unity for the first injection until a consistent value was reached that was approximately one half to one third of the maximum value. A comparable trend was observed for mass delivered per injection. For the conditions used in this investigation, the source efficiency decreased when the injection time was increased to longer than 1 s, when pressure was decreased, and when the carrier gas flow rate was increased. Although the corresponding mass per injection increased with these changes, the increase in mass was less than that predicted had the carrier gas been saturated. The source efficiency did not depend strongly on temperature and only moderately on vessel idle durations (4–16 s). The degree of PDMAT decomposition was evaluated by measuring the partial pressure of dimethylamine (the primary PDMAT decomposition product under the conditions of this investigation) using the same gas analyzer. For a given idle time, the amount of dimethylamine delivered more than doubled as vessel temperature was increased from 68 to 78 °C.

Thin film vapor deposition processes, e.g., chemical vapor deposition, are widely used in high-volume manufacturing of electronic and optoelectronic devices. Ensuring desired film properties and maximizing process yields require control of the chemical precursor flux to the deposition surface. However, achieving the desired control can be difficult due to numerous factors, including delivery system design, ampoule configuration, and precursor properties. This report describes an apparatus designed to investigate such factors. The apparatus simulates a single precursor delivery line, e.g., in a chemical vapor deposition tool, with flow control, pressure monitoring, and a precursor-containing ampoule. It also incorporates an optical flow cell downstream of the ampoule to permit optical measurements of precursor density in the gas stream. From such measurements, the precursor flow rate can be determined, and, for selected conditions, the precursor partial pressure in the headspace can be estimated. These capabilities permit this apparatus to be used for investigating a variety of factors that affect delivery processes. The methods of determining the pressure to (1) calculate the precursor flow rate and (2) estimate the headspace pressure are discussed, as are some of the errors associated with these methods. While this apparatus can be used under a variety of conditions and configurations relevant to deposition processes, the emphasis here is on low-volatility precursors that are delivered at total pressures less than about 13 kPa downstream of the ampoule. An important goal of this work is to provide data that could facilitate both deposition process optimization and ampoule design refinement.

High volume manufacturing of devices based on transition metal dichalcogenide (TMD) ultra-thin films will require deposition techniques that are capable of reproducible wafer-scale growth with monolayer control. To date, TMD growth efforts have largely relied upon sublimation and transport of solid precursors with minimal control over vapor phase flux and gas-phase chemistry, which are critical for scaling up laboratory processes to manufacturing settings. To address these issues, we report a new pulsed metalorganic chemical vapor deposition (MOCVD) route for MoS2 film growth in a research-grade single-wafer reactor. Using bis(tert-butylimido)-bis(dimethylamido)molybdenum and diethyl disulfide we deposit MoS2 films from ≈ 1 nm to ≈ 25 nm in thickness on SiO2/Si substrates. We show that layered 2H-MoS2 can be produced at comparatively low reaction temperatures of 591 °C at short deposition times, approximately 90 s for few-layer films. In addition to the growth studies performed on SiO2/Si, films with wafer-level uniformity are demonstrated on 50 mm quartz wafers. Process chemistry and impurity incorporation from precursors are also discussed. This low-temperature and fast process highlights the opportunities presented by metalorganic reagents in the controlled synthesis of TMDs.

The design and operation of a simple, optically-accessible modular reactor for probing thermal thin film deposition processes, such as atomic layer deposition processes (ALD) and chemical vapor deposition (CVD), is described. This reactor has a nominal footprint of 225 cm(2) and a mass of approximately 6.6 kg, making it small enough to conveniently function as a modular component of an optical train. The design is simple, making fabrication straightforward and relatively inexpensive. Reactor operation is characterized using two infrared absorption measurements to determine exhaust times for tetrakis(dimethylamino)titanium and water, proto-typical ALD precursors, in a pressure and flow regime commonly used for ALD.

The performance of a bubbler to deliver the low-volatility, liquid cobalt precursor μ2-η2-(tBu-acetylene) dicobalthexacarbonyl (CCTBA) for reduced-pressure chemical vapor deposition and atomic layer deposition processes was characterized. A relatively large process window was investigated by varying carrier gas flow rate, system pressure, and bubbler temperature. For this range of conditions, the CCTBA partial pressure was measured using a custom-designed nondispersive infrared gas analyzer, and the CCTBA flow rates were derived from these partial pressure measurements. The dependence of CCTBA flow rate on these process parameters was modeled to obtain a deeper understanding of the factors influencing bubbler performance. Good agreement between measured and modeled CCTBA flow rates was obtained using a model in which the pressure drop between the bubbler head space and the pressure measurement location was included and in which a constant CCTBA partial pressure in the bubbler head space for a given bubbler temperature was assumed. The dependence of CCTBA head space partial pressure on temperature was parameterized in the form of the August equation, which is commonly used to describe the temperature-dependence of vapor pressure. While this report was focused specifically on CCTBA, the results of this study indicate that this method for estimating the precursor delivery rate from a bubbler should be applicable to other low-volatility, liquid precursors.The performance of a bubbler to deliver the low-volatility, liquid cobalt precursor μ2-η2-(tBu-acetylene) dicobalthexacarbonyl (CCTBA) for reduced-pressure chemical vapor deposition and atomic layer deposition processes was characterized. A relatively large process window was investigated by varying carrier gas flow rate, system pressure, and bubbler temperature. For this range of conditions, the CCTBA partial pressure was measured using a custom-designed nondispersive infrared gas analyzer, and the CCTBA flow rates were derived from these partial pressure measurements. The dependence of CCTBA flow rate on these process parameters was modeled to obtain a deeper understanding of the factors influencing bubbler performance. Good agreement between measured and modeled CCTBA flow rates was obtained using a model in which the pressure drop between the bubbler head space and the pressure measurement location was included and in which a constant CCTBA partial pressure in the bubbler head space for a given bubble...

In situ infrared (IR) spectroscopy has proven to be an extremely valuable tool for understanding various aspects of chemical vapor deposition (CVD) and atomic layer deposition (ALD). Solid film deposits, gaseous precursors and by-products, and growth surfaces are all of interest in these processes, and IR spectroscopy is capable of providing useful data for each. Qualitative measurements are sometimes sufficient, but quantitative spectroscopy is often desired for work involving process monitoring or process development. Relating absorbancetoconcentrationscanbeproblematicforsurfacesandsolid deposits due to interactions, dielectric screening, etc., but measurements of the gas phase are straightforward if absorptivities are known. Efforts to obtain accurate IR reference spectra of CVD and ALD precursors, however, have been limited. 1 Commercial quantitative spectral libraries, if they include any precursors at all, typically contain only the most common ones (e.g., SiH4 and WF6). This is despite the use of IR spectroscopy for process monitoring in manufacturing 2 and for process development in the laboratory. 3,4 There is a need, therefore, for carefully obtained spectra of the less common precursors and for a general method of obtaining them. This work presents a static measurement for determining the IR absorptivities of low vapor pressure organometallic compounds used as precursors for CVD and ALD. We use Fourier transform IR (FTIR) spectroscopy to this end. Since the focus of this work is related to CVD and ALD processes where the compounds are typically heated to increase their vapor pressure, we obtain reference spectra at elevated temperatures. Heated samples are known to cause photometric errors in FT-IR spectroscopy, so we use a setup that is demonstrated to provide correct spectra. The apparatus and method for emission correction are discussed fully in the Experimental section. Quantitatively accurate IR spectra of tetrakis(dimethylamido)titanium (TDMAT), Ti[N(CH3)2]4 ,a re measured in the temperature range of (352 to 476) K. TDMAT is an alkylamido organometallic compound, a class of molecules that has found widespread use as CVD and ALD precursors in the semiconductor industry. TDMAT, in particular, is commonly used for the CVD of TiN and is under investigation for the ALD of TiO2. Qualitative IR spectra of TDMAT have been reported a number of times in the literature, 5‐11 which offers an opportunity to compare the results to previous studies and to elucidate the discrepancies. The temperature-dependent IR absorptivity is presented in the Results section, and a comparison to prior results can be found in the ∗

This work presents a novel method for obtaining surface infrared spectra with sub-second time resolution during atomic layer deposition (ALD). Using a rapid-scan Fourier transform infrared (FT-IR) spectrometer, we obtain a series of synchronized interferograms (120 ms) during multiple ALD cycles to observe the dynamics of an average ALD cycle. We use a buried metal layer (BML) substrate to enhance absorption by the surface species. The surface selection rules of the BML allow us to determine the contribution from the substrate surface as opposed to that from gas-phase molecules and species adsorbed at the windows. In addition, we use simulation to examine the origins of increased reflectivity associated with phonon absorption by the oxide layers. The simulations are also used to determine the decay in enhancement by the buried metal layer substrate as the oxide layer grows during the experiment. These calculations are used to estimate the optimal number of ALD cycles for our experimental method.

Metal alkylamide compounds, such as tetrakis(ethylmethylamido) hafnium (TEMAH), represent a technologically important class of metalorganic precursors for the deposition of metal oxides and metal nitrides via atomic layer deposition (ALD) or chemical vapor deposition. The development of in situ diagnostics for processes involving these compounds could be beneficial in, e.g., developing deposition recipes and validating equipment-scale simulations. This report describes the performance of the combination of two techniques for the simultaneous, rapid measurement of the three major gas phase species during hafnium oxide thermal ALD using TEMAH and water: TEMAH, water, and methylethyl amine (MEA), the only major reaction by-product. For measurement of TEMAH and MEA, direct absorption methods based on a broadband infrared source with different mid-IR bandpass filters and utilizing amplitude modulation and synchronous detection were developed. For the measurement of water, wavelength modulation spectroscopy utilizing a near-IR distributed feedback diode laser was used. Despite the relatively simple reactor geometry employed here (a flow tube), differences were easily observed in the time-dependent species distributions in 300 mL/min of a helium carrier gas and in 1000 mL/min of a nitrogen carrier gas. The degree of TEMAH entrainment was lower in 300 mL/min of helium compared to that in 1000 mL/min of nitrogen. The capability to obtain detailed time-dependent species concentrations during ALD could potentially allow for the selection of carrier gas composition and flow rates that would minimize parasitic wall reactions. However, when nitrogen was employed at the higher flow rates, various flow effects were observed that, if detrimental to a deposition process, would effectively limit the upper range of useful flow rates.

In this work, a Fourier transform infrared spectroscopy-based method is developed to measure the gas-phase dynamics occurring during atomic layer deposition. This new technique is demonstrated during the deposition of hafnium oxide using tetrakis(ethylmethylamido)hafnium and water vapor. The repeatability of the deposition process is utilized to signal average across multiple cycles. This approach required synchronizing the precursor injection pulses with the moving mirror of the spectrometer. The system as implemented in this work achieves spectra with a time resolution of ≈150 ms, but better resolution can be easily obtained. Using this technique, the authors are able to optically measure transients in the molecular number densities of the precursors and product that are the effects of mass transport and surface reactions.

Tetrakis(ethylmethylamido)hafnium and water are commonly used precursors for atomic layer deposition of HfO 2 . Using reflection absorption infrared spectroscopy with a buried-metal-layer substrate, we probe surface species present during typical deposition conditions. We observe evidence for thermal decomposition of alkylamido ligands at 320 °C. Additionally, we find that complete saturation of the SiO 2 substrate occurs in the first cycle at ≈100 °C whereas incomplete coverage is apparent even after many cycles at higher temperatures. The use of this technique as an in situ diagnostic useful for process optimization is demonstrated.

Internal photoemission (IPE) spectroscopy is a powerful technique for investigating electronic properties at solid-solid interfaces. Upon photon excitation, electrons or/and holes in the solid under an external electrical bias, accumulate at the interface, escape over the energetic interfacial barrier, and produce an externally measured photocurrent. The onset of the photocurrent as a function of photon excitation energy indicates the energy barrier height as commonly used in the threshold spectroscopy. Furthermore, the characteristics of photoemission yield at the energies above (or possibly below) the barrier can also yield spectroscopic information. IPE has recently become an important tool used to determine the band offsets in metal oxide semiconductor structures. In this report, we will present a brief description of IPE principles and experimental setup and apply IPE to two important GaAs-based metal oxide semiconductor structures. These structures have become of prime research interest as potential structures in the next generation of electronic devices when the current silicon-based device downscaling comes to a halt.

Atomic layer deposition (ALD) is an important method for depositing the nanometer-scale, conformal high k dielectric layers required for many nanoelectronics applications. In situ monitoring of ALD processes has the potential to yield insights that will enable efficiencies in film growth, in the development of deposition recipes, and in the design and qualification of reactors. This report will describe the status of a project to develop in situ diagnostics for hafnium oxide ALD processes. The focus is on an examination of the utility of Fourier transform infrared spectroscopy and diode laser spectroscopy for optimizing deposition conditions, rather than simply monitoring precursor delivery. Measurements were performed in a single-wafer, warm-wall, horizontal-flow reactor during hafnium oxide ALD involving tetrakis(ethylmethylamino) hafnium and water. Measurements were performed near the wafer surface under a range of deposition conditions in an effort to correlate gas phase measurements with surface processes.

In this work, the species present in the gas phase during atomic layer deposition of hafnium oxide were investigated in an attempt to gain insight into the chemistry of this system. Hafnium oxide was deposited on a silicon substrate using tetrakis(ethylmethylamino) hafnium, Hf[N(C2H5)(CH3)]4, and water. In situ infrared absorption spectroscopy measurements were performed in a research-grade, horizontal-flow reactor under a range of deposition conditions. Density functional theory quantum calculations of vibrational frequencies of expected species were used to facilitate identification of observed spectral features.

The effects of substrate temperature and near-substrate plasma density are studied for reactively sputtered dc magnetron aluminum doped zinc oxide (ZnO:Al). Plasma density is varied using an unbalanced magnetron along with external Helmholtz coils, and is characterized using flat and cylindrical Langmuir probes. The substrate ion-to-neutral flux ratio was varied from 0.2 to about 3.5 using this technique. The ZnO:Al films were characterized by resistivity, transmission, Hall effect, and theta-two theta x-ray diffraction. Under conditions of low plasma density, for substrate temperatures below 65 °C the substrate temperature does not have a significant influence on film quality, and the film quality is relatively poor. By 88 °C the film quality is rapidly improving with temperature, and at 135 °C high quality films are produced. For substrate temperatures below 65 °C, increased plasma density at the substrate has a significant beneficial effect on the optoelectronic quality of the films. At 105 °C this eff...

In this paper, we have fabricated Bi2Se3 nanowire FETs by using a self-alignment technique and observed excellent device characteristics. The FETs show unipolar, n-type behavior with a clear cutoff in the OFF-state with only thermally activated conduction at relatively high temperatures, and a well-saturated output current indicating surface metallic conduction. These data illustrate that charge transport in materials associated with topological insulator (TI) systems may be much more complicated than the conventional wisdom that has recently been developed for these novel systems.

Topological insulators are unique electronic materials with insulating interiors and robust metallic surfaces. Device applications exploiting their remarkable properties have so far been hampered by the difficulty to electrically tune the Fermi levels of both bulk and thin film samples. Here we show experimentally that single-crystal nanowires of the topological insulator Bi2Se3 can be used as the conduction channel in high-performance field effect transistor (FET), a basic circuit building block. Its current-voltage characteristics are superior to many of those reported for semiconductor nanowire transistors, including sharp turn-on, nearly zero cutoff current, very large On/Off current ratio, and well-saturated output current. The metallic electron transport at the surface with good FET effective mobility can be effectively separated from the conduction of bulk Bi2Se3 and adjusted by field effect at a small gate voltage. This opens up a suite of potential applications in nanoelectronics and spintronics.

Water is often employed as the oxygen source in metal oxide atomic layer deposition (ALD) processes. It has been reported that variations in the amount of water delivered during metal oxide ALD can impact the oxide film properties. Hence, one contribution to optimizing metal oxide ALD processes would be to identify methods to better control water dose. The development of rapid, quantitative techniques for in situ water vapor measurements during ALD processes would be beneficial to achieve this goal. In this report, the performance of an in situ tunable diode laser absorption spectroscopy (TDLAS) scheme for performing rapid, quantitative water partial pressure measurements in a representative quarter-inch ALD delivery line is described. This implementation of TDLAS, which utilizes a near-infrared distributed-feedback diode laser and wavelength modulation spectroscopy, provides measurements of water partial pressure on a timescale comparable to or shorter than the timescale of the gas dynamics in typical AL...

A description is given of the design and performance of a diagnostic-accessible, perpendicular-flow, single-wafer deposition reactor for use with 50 mm wafers. The reactor chamber design is based on a simple flow tube, with diagnostic access achieved by replacing sections of the reactor chamber wall with recessed diagnostic ports. Reactor chamber performance is evaluated for the purpose of performing optical measurements during atomic layer deposition (ALD). Computational fluid dynamics simulations predict that the when used with windows the diagnostic port design produces minimal perturbations to the gas flow under typical deposition conditions, as compared to a design without diagnostic ports. Temperature measurements of the inside surface of a window installed in a diagnostic port suggest that for reactor chamber operation at 110 °C, under typical deposition conditions, the inside surface window temperature is approximately equal to or greater than the surrounding reactor chamber temperature, thereby minimizing possible species condensation on the window surface. As a consequence of using recessed diagnostic ports, an increase in the amplitude of optical intensity fluctuations was generally observed when performing measurements at elevated chamber temperatures. These intensity fluctuations could be readily reduced by enclosing the optical path to the exterior side of the windows. The performance of two straight-forward methods to reduce these intensity fluctuations is presented. The results outlined above demonstrate that this reactor design can be operated with short gas residence times and with all reactor surfaces at elevated temperatures, making it useful for simulating a wide range of gas flow conditions with relevance to microelectronics-related ALD processes.

An in situ gas-phase diagnostic for the metal alkylamide compound tetrakis(ethylmethylamido) hafnium (TEMAH), Hf[N(C2H5)(CH3)]4, was demonstrated. This diagnostic is based on direct absorption measurement of TEMAH vapor using an external cavity quantum cascade laser emitting at 979 cm−1, coinciding with the most intense TEMAH absorption in the mid-infrared spectral region, and employing 50 kHz amplitude modulation with synchronous detection. Measurements were performed in a single-pass configuration in a research-grade atomic layer deposition (ALD) chamber. To examine the detection limit of this technique for use as a TEMAH delivery monitor, this technique was demonstrated in the absence of any other deposition reactants or products, and to examine the selectivity of this technique in the presence of deposition products that potentially interfere with detection of TEMAH vapor, it was demonstrated during ALD of hafnium oxide using TEMAH and water. This technique successfully detected TEMAH at molecular densities present during simulated industrial ALD conditions. During hafnium oxide ALD using TEMAH and water, absorbance from gas-phase reaction products did not interfere with TEMAH measurements while absorption by reaction products deposited on the optical windows did interfere, although interfering absorption by deposited reaction products corresponded to only ≈4% of the total derived TEMAH density. With short measurement times and appropriate signal averaging, estimated TEMAH minimum detectable densities as low as ≈2 × 1012 molecules/cm3 could be obtained. While this technique was demonstrated specifically for TEMAH delivery and hafnium oxide ALD using TEMAH and water, it should be readily applicable to other metal alkylamide compounds and associated metal oxide and nitride deposition chemistries, assuming similar metal alkylamide molar absorptivity and molecular density in the measurement chamber.

In this work, measurements were performed in a single-wafer, warm-wall, horizontal-flow reactor. Additional design characteristics of this reactor include optical access near the wafer surface, good gas flow characteristics (facilitating reproducible high quality film growth), and an aluminum body (facilitating maintenance of a uniform wall temperature).

The advent of the cable-less lightpipe radiation thermometer (CLRT) has resulted in a significant improvement in the accuracy of CLRT calibrations and measurements. CLRT systems show great promise in noncontact measurements by the elimination of the uncertainties caused by the long fiber optic cables and their connections and by the extension of the spectral range to handle low temperature applications down to ambient conditions. Calibration of the CLRT at the National Institute of Standards and Technology (NIST) will be performed with the water heat pipe blackbody source between 50 °C and 230 °C. In addition, the CLRT will be compared to contact thermometers on a silicon wafer heated in a Post-Exposure Bake test bed at NIST. Comparison of the CLRT with both the blackbody and thermocouple standards will provide confidence in using CLRTs and will allow researchers to continued research in improving the accuracy and feasibility of applying CLRTs in semiconductor processing.

Since pyrometric thermometry is a noncontact method, it has great promise as a technique for monitoring silicon wafers during rapid thermal processing (RTP). Absolute values of surface emissivity are required when making pyrometric temperature measurements. One approach to obtaining these values is the use of emissivity compensated pyrometry, where a reflectometer is integrated into the pyrometer to allow real-time emissivity measurement. While this technique has been successfully applied to metal organic chemical vapor deposition (MOCVD) of compound semiconductors, it has not been applied to RTP. Although such measurements require that the surface be a specular reflector, they promise real-time traceable temperature measurements that are independent of the nature of the wafer. Here we discuss measurement of wafer temperature for polished wafers and an initial attempt to measure a patterned wafer during heating inside the RTP test bed at the National Institute of Standards and Technology

We report on the status of a combined experimental and modeling study for atomic layer deposition (ALD) of HfO2 and Al2O3. Hafnium oxide films were deposited from tetrakis(dimethylamino)hafnium and water. Aluminum oxide films from trimethyl aluminum and water are being studied through simulations. In this work, both in situ metrologies and process models are being developed. Optically‐accessible ALD reactors have been constructed for in situ, high‐sensitivity Raman and infrared absorption spectroscopic measurements to monitor gas phase and surface species. A numerical model using computational fluid dynamics codes has been developed to simulate the gas flow and temperature profiles in the experimental reactor. Detailed chemical kinetic models are being developed with assistance from quantum chemical calculations to explore reaction pathways and energetics. This chemistry is then incorporated into the overall reactor models.

Lightpipe radiation thermometers (LPRTs) are used as temperature monitoring sensors in rapid thermal processing (RTP) tools for semiconductor fabrication. In order to assure uniform wafer temperatures during processing these RTP tools generally have highly reflecting chamber walls to promote a uniform heat flux on the wafer. To minimize disturbances in the chamber reflectivity small, 2 mm diameter, sapphire lightpipes are often the temperature sensor of choice. This study was undertaken to measure and model the effect of LPRT proximity on the wafer temperature. Our experiments were performed in the NIST RTP test bed. We measured the spectral radiance temperature with the center lightpipe and compared these with the three LPRTs at the mid-radius of the wafer and the thin-film thermocouple (TFTC) junctions of a NIST calibration wafer. Depressions in the wafer temperature up to 25/spl deg/C with the lightpipe at 2 mm spacing were observed. A finite-element radiation model of the wafer-chamber-lightpipe was developed to predict the temperature depression as a function of proximity distance and separation distance. The experimental results were compared with those from a model that accounts for lightpipe geometry and radiative properties, wafer emissivity and chamber cold plate reflectivity.

Recent studies on dynamic temperature profiling and lithographic performance modeling of the post-exposure bake (PEB) process have demonstrated that the rate of heating and cooling may have an important influence on resist lithographic response. Measuring the transient surface temperature during the heating or cooling process with such accuracy can only be assured if the sensors embedded in or attached to the test wafer do not affect the temperature distribution in the bare wafer. In this paper we report on an experimental and analytical study to compare the transient response of embedded platinum resistance thermometer (PRT) sensors with surface-deposited, thin-film thermocouples (TFTC). The TFTCs on silicon wafers have been developed at NIST to measure wafer temperatures in other semiconductor thermal processes. Experiments are performed on a test bed built from a commercial, fab-qualified module with hot and chill plates using wafers that have been instrumented with calibrated type-E (NiCr/CuNi) TFTCs and commercial PRTs. Time constants were determined from an energy-balance analysis fitting the temperature-time derivative to the wafer temperature during the heating and cooling processes. The time constants for instrumented wafers ranged from 4.6 s to 5.1 s on heating for both the TFTC and PRT sensors, with an average difference less than 0.1 s between the TFTCs and PRTs and slightly greater differences on cooling.

Lightpipe radiation thermometers (LPRTs) are widely used to measure wafer temperatures in rapid thermal processing (RTP) tools. To use blackbody‐calibrated LPRTs to infer the wafer temperature, it is necessary to build a model to predict the effective emissivity accounting for the wafer and chamber radiative properties as well as geometrical features of the chamber. The uncertainty associated with model‐corrected temperatures can be investigated using test wafers instrumented with thin‐film thermocouples (TFTCs) on which the LPRT target spot has been coated with films of different emissivity. A finite‐element model of the wafer‐chamber arrangement was used to investigate the effects of Pt spot (es = 0.25) and Au spot (es = 0.05) on the temperature distribution of test wafers with spectral emissivities of 0.65 and 0.84. The effects of the shield reflectivity and the cool lightpipe (LP) tip on the wafer temperature were evaluated. A radiance analysis method was developed, and a comparison of model‐based predictions with experimental observations was made on a 200 mm diameter wafer in the NIST RTP test bed. The temperature rises caused by the low‐emissivity spot were predicted and the cooling effect of the LP tip was determined. The results of the study are important for developing the model‐based corrections for temperature measurements and related uncertainties using LPRTs in semiconductor thermal processes.

Lightpipe radiation thermometers (LPRTs) are used as temperature monitoring sensors in most rapid thermal processing (RTP) tools for semiconductor fabrication. These tools are used for dopant anneal, gate oxide formation, and other high temperature processing. In order to assure uniform wafer temperatures during processing these RTP tools generally have highly reflecting chamber walls to promote a uniform heat flux on the wafer. Therefore, only minimal disturbances in the chamber reflectivity are permitted for the sensors, and the small 2 mm diameter sapphire lightpipe is generally the temperature sensor of choice. This study was undertaken to measure and model the effect of LPRT proximity on the wafer temperature. Our experiments were performed in the NIST RTP test bed using a NIST thin‐film thermocouple (TFTC) calibration wafer. We measured the spectral radiance temperature with the center lightpipe and compared these with the TFTC junctions and with the three LPRTs at the mid‐radius of the wafer. We me...

Rapid thermal processing (RTP) tools are currently monitored and controlled with lightpipe radiation thermometers (LPRTs) which have been calibrated with thermocouple instrumented wafers. We have developed a thin‐film thermocouple wafer that enables more accurate calibration of the LPRTs. The NIST thin‐film thermocouple calibration wafer uses Pt/Pd wire thermocouples welded to thin‐film Rh/Pt thermocouples to reduce the uncertainty of the wafer temperature measurement in situ. We present the results of testing these thin‐film thermocouple calibration wafers in the NIST RTP test bed at temperatures ranging from 650 °C to 830 °C together with a discussion of the materials limitations and capabilities. The difference between the thermocouple junction temperatures and the radiance temperatures indicated by the blackbody‐calibrated LPRT can be attributed to the effective emissivity of the wafer, the parameter that accounts for the geometry and radiative properties of the wafer‐chamber configuration. An analysis of the uncertainty, u = 1.3 K (k =1), of the wafer surface temperature measurements in the NIST RTP test bed is presented. Confirmation of this value was partially hampered by thermal gradients in the chamber and some problems with the weld pads at high temperature. In addition, we discuss the determination of the Seebeck coefficient of the thin‐film thermocouples used on the wafers.

Atomic layer deposition of titanium dioxide using tetrakis(dimethylamido)titanium (TDMAT) and water vapor is studied by reflection-absorption infrared spectroscopy (RAIRS) with a time resolution of 120 ms. At 190 °C and 240 °C, a decrease in the absorption from adsorbed TDMAT is observed without any evidence of an adsorbed product. Ex situ measurements indicate that this behavior is not associated with an increase in the impurity concentration or a dramatic change in the growth rate. A desorbing decomposition product is consistent with these observations. RAIRS also indicates that dehydroxylation of the growth surface occurs only among one type of surface hydroxyl groups. Molecular water is observed to remain on the surface and participates in reactions even at a relatively high temperature (110 °C) and with long purge times (30 s).

We have fabricated Pt thin-film resistors directly sputtered on silicon substrates to evaluate their use as resistance thermal detectors (RTDs). This technique was chosen to achieve more accurate temperature measurements of large silicon wafers during semiconductor processing. High-purity (0.999 968 mass fraction) platinum was sputter deposited on silicon test coupons using titanium and zirconium bond coats. These test coupons were annealed, and four-point resistance specimens were prepared for thermal evaluation. Their response was compared with calibrated platinum–palladium thermocouples in a tube furnace. We evaluated the effects of furnace atmosphere, thin-film thickness, bond coats, annealing temperature and peak thermal excursion of the Pt thin films. Secondary ion mass spectrometry (SIMS) was performed to evaluate the effect of impurities on the thermal resistance coefficient, α. We present typical resistance versus temperature curves, hysteresis plots versus temperature and an analysis of the causes of uncertainties in the measurement of seven test coupons. We conclude that sputtered thin-film platinum resistors on silicon wafers can yield temperature measurements with uncertainties of less than 1 °C, k = 1 up to 600 °C. This is comparable to or better than commercially available techniques.

We present a non-contact method for the determination of the thermal response time of temperature sensors embedded in wafers. In this method, a flash lamp illuminates a spot on the wafer in periodic pulses; the spot is on the opposite side from the sensor under test. The thermal time constant of the sensor is then obtained from measurement of its temporal response, together with a theoretical model of heat flows both into the sensor and laterally within the wafer. Experimental data on both platinum resistance thermometers (PRTs) and on thermocouples embedded in silicon wafers show good agreement with the heat transfer models. Values of the thermal response time for a wide range of experimental parameters agree to within standard deviations of 8% (PRTs) and 20% (thermocouples), demonstrating the self-consistency of our results. The method is directly applicable to determining the thermal properties of sensors used in instrumented silicon wafers. We anticipate that the method will have use in development of new sensor attachment methods, in verifying the proper attachment of sensors during production, and in confirming that the thermal attachment has not degraded with age or thermal cycling. To simplify the application of the method, we have produced a table of calculated relevant quantities to be used in relating the measured signal to the thermal response time.