Classifications

• • H—ELECTRICITY
  • H10—SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
  • H10P—GENERIC PROCESSES OR APPARATUS FOR THE MANUFACTURE OR TREATMENT OF DEVICES COVERED BY CLASS H10
  • H10P34/00—Irradiation with electromagnetic or particle radiation of wafers, substrates or parts of devices
  • H10P34/40—Irradiation with electromagnetic or particle radiation of wafers, substrates or parts of devices with high-energy radiation
  • H10P34/42—Irradiation with electromagnetic or particle radiation of wafers, substrates or parts of devices with high-energy radiation with electromagnetic radiation, e.g. laser annealing
• • H—ELECTRICITY
  • H10—SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
  • H10P—GENERIC PROCESSES OR APPARATUS FOR THE MANUFACTURE OR TREATMENT OF DEVICES COVERED BY CLASS H10
  • H10P14/00—Formation of materials, e.g. in the shape of layers or pillars
  • H10P14/20—Formation of materials, e.g. in the shape of layers or pillars of semiconductor materials
• • H—ELECTRICITY
  • H10—SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
  • H10P—GENERIC PROCESSES OR APPARATUS FOR THE MANUFACTURE OR TREATMENT OF DEVICES COVERED BY CLASS H10
  • H10P14/00—Formation of materials, e.g. in the shape of layers or pillars
  • H10P14/20—Formation of materials, e.g. in the shape of layers or pillars of semiconductor materials
  • H10P14/38—Formation of materials, e.g. in the shape of layers or pillars of semiconductor materials characterised by treatments done after the formation of the materials
  • H10P14/3802—Crystallisation or recrystallisation of non-monocrystalline semiconductor materials, e.g. regrowth
  • H10P14/3808—Crystallisation or recrystallisation of non-monocrystalline semiconductor materials, e.g. regrowth using laser beams
• • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
  • Y10—TECHNICAL SUBJECTS COVERED BY FORMER USPC
  • Y10S—TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
  • Y10S148/00—Metal treatment
  • Y10S148/09—Laser anneal
• • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
  • Y10—TECHNICAL SUBJECTS COVERED BY FORMER USPC
  • Y10S—TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
  • Y10S148/00—Metal treatment
  • Y10S148/093—Laser beam treatment in general
• • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
  • Y10—TECHNICAL SUBJECTS COVERED BY FORMER USPC
  • Y10S—TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
  • Y10S438/00—Semiconductor device manufacturing: process
  • Y10S438/965—Shaped junction formation

Definitions

• laser or electron beam annealing is a new processing tool, not only for annealing ion damage effectively in very short time periods, but for relocating impurities in semiconductors on a selective basis.
• the light reflectivity of the surface of the material is monitored during.the annealing process. After the annealing beam is turned on, the surface melts, and the reflectivity of the surface changes. It changes again when the surface solidifies sometime after the beam is turned off. The time difference between these changes is the lifetime of the molten surface. Precise determination of the extent of annealing is obtained in response to the duration of the molten zone lifetime. Control can be effected using this information.
• FIG. 1 is a schematic illustration of the impurity redistribution effect occurring during laser annealing; •
• FIG. 2 is a schematic representation of an apparatus useful for effecting the result described by FIG. 1.
• FIG. 3 is a plot showing, as a function of time, t, the intensity of the laser annealing pulse, I Q , the transmitted laser intensity, I t (in arbitrary units), and the absolute value of the reflectivity R of the surface on which the pulse is incident;
• FIG. 4 is a plot of the lifetime of the molten zone as a function of the square of the incident laser energy per unit area E;
• FIG. 5 is a plot showing the initial and final impurity distributions for an exemplary laser anneal
• FIG. 6 is a plot of the final depth of an impurity layer as a function of the measured lifetime of the molten region.
• FIG. 1 shows a substrate material 10, which in a preferred embodiment is a semiconductor such as silicon or gallium arsenide, or other material in which
• a masking layer 11 e.g.,
• the laser annealing energy indicated by the schematic beam 13 is incident on the window.
• Numeral 12 indicates the initial position of the impurity boundary after (in this embodiment) ion implantation (and in this case before anneal) .
• the portion illustrated, the boundary of which is denoted by line 14 becomes molten.
• the impurities diffuse in the liquid phase, forming an ultimate region with a boundary designated 15.
• the annealing process can be adapted and controlled either to minimize redistribution or to achieve a controlled movement of impurities.
• the existence and duration of the molten region is detected by directing a beam from laser 20 or other appropriate radiant energy beam, onto the surface exposed to the beam from the annealing laser 21.
• a detector 22 for measuring the intensity of the light reflected from that surface is placed so as to receive the reflected beam.
• a detector 23, situated to receive energy from the annealing laser 21 penetrating the wafer ("transmitted" energy) was also used to obtain the data presented below, but is not necessary for the practice of the invention.
• a timing circuit 24 is connected as a feedback loop between the detector 22 and the laser 21 as shown. Its use comprises a preferred embodiment in which the extent of laser annealing or the control of the depth of the impurity region is in automatic response to the measured lifetime of the molten region.
• a Q-switched Nd:glass oscillator and amplifier system was used to generate the annealing beam at a wavelength of 1.06 ⁇ m.
• Mode-control was used in the oscillator to produce a single transverse mode having a Gaussian profile and a diameter (FWHM) of 1.2 mm at the oscillator output and 3.3 mm at the samples.
• the beam is generated as a series of pulses, the duration of the pulses being 50 nsec (FWHM) .
• the annealing is accomplished by the application of a single pulse to the substrate.
• a continuous 0.63 ⁇ m HeNe laser was used to measure the optical reflectivity of the implanted surface during the annealing. This beam was focused to a diameter of approximately 0.3 mm and was aligned with the center of the annealing beam at the crystal surface. The angle of incidence of the 0.63 ⁇ m beam was 57.5 degrees to the surface normal.
• a polarizer and quarter wave plate were employed to measure the reflectivity for both perpendicular and parallel polarizations, R ⁇ and i, respectively.
• OMPl approximately 5 nsec was used to detect the reflected beam. Accurate timing was insured by synchronously triggering the recording of the incident and transmitted 1.06 ⁇ m pulses and the 0.63 ⁇ m reflectivity signal.
• the extent of the annealing in the samples was determined by Rutherford backscattering and channeling analysis using 1.9 meV alpha particles. For an incident laser energy greater than 1.8 _+ 0.3 J/cm 2 in the 10-"-° ions/cm sample, the channelled spectrum of the annealed sample, had a minimum yield of approximately 3 percent for both the Si and As spectra. This value is comparable to that of uni planted single-crystal Si.
• the reflectivity measurements are illustrated in FIG. 3. These waveforms have a number of distinguishing features which constitute characteristic "signature" for laser annealing.
• a laser beam pulse I Q is generated, the pulse, for the particular laser source used (other sources being usable) , having the rise and fall time shown.
• Some of the laser energy penetrates the sample and is illustarted as the transmitted energy pulse I t .
• a sharp transition occurred when the transmitted 1.06 ⁇ m energy I t decreased abruptly, and simultaneously the reflectivity of the monitoring 0.63 ⁇ m beam increased from an initial value R a to a new value R -.
• R l - R? were 0.63, 0.82 and 0.57
• the onset of reflectivity change at time t (FIG. 3) was observed to move back towards the onset edge of the incident annealing pulse as the pulse energy increased.
• the integrated energy of the incident laser pulse from the beginning of the pulse to the time t j _ of reflectivity change was found to be nearly constant for a given sample over the measured range of total pulse energies and found to be 0.9 +; .1 J/cm" 1 and 1.3 + .05 J/cm 2 for 10 16 and 10 15 ions/cm 2 samples, respectively.
• OMPI WIPO decrease in transmission of the annealing beam at time t j L when the surface melts.
• the constant value of the reflectivity between t j _ and t 2 is strong evidence that the surface remains a liquid during this time interval.
• the absorption of the annealing beam is determined entirely by the optical properties of the liquid, and the beam energy is deposited at a thin layer at the surface.
• the penetration depth will be approximately ⁇ 2D L ⁇ , where ⁇ is the time difference between t 2 and t ⁇ , and the diffusion coefficient of As in liquid silicon is estimated to be 1 x 10 ⁇ 4 cm 2 s ⁇ .
• the As will then diffuse approximately 1300 Angstroms. Since the initial distribution was peaked at 320 Angstroms, the final distribution would be expected to be 1/e of its maximum experimental value of 2000 Angstroms.
• the duration of the "flat-top" portion of the reflectivity signature corresponds to the time required for the complete solidification of the liquid layer at the crystal surface. As illustrated in FIG. 4, this time varied as the square of the incident energy.
• the time ⁇ is the time required for the absorbed energy per unit area, E a , to diffuse into the solid so that it is stored in the heat capacity of hot solid (i.e., at T ⁇ T M ) .
• Equation (1) predicts the 5 observed dependence of ⁇ on E. Taking into account the reflectivity of the liquid surface at 1.06 ⁇ m, E a is 0.23 E. Using Equation (1), we then estimate the slope of the
• the 20 layer could also be formed by implantation of "inert” ions such as helium, or argon.
• the layer could be evaporated or otherwise deposited onto a single crystal substrate either of the same material or a material with a matching lattice, thus forming a
• redistribution could be done as a separate step, or as an error correction or healing measure.
• redistribution of impurities in metals, amorphous materials, or polycrystalline materials can be
• the correlation between the depth to which the impurity region redistributes and the lifetime of the molten zone can be made empirically or can be calculated using liquid-phase diffusion constants. The former are possible using experiments such as that just described. The calculations can follow those just described although the liquid state diffusion constants are not as well established as those for the solid state.
• FIG. 6 A typical correlation curve based on the data described earlier is shown in FIG. 6.
• Greatest depth re ⁇ distribution occurs when the lifetime of the molten zone is longest, thus indicating reflects penetration of the zone to below the initial location of impurities.
• the redistribution is small where the depth of the molten layer approximates the impurity depth as is the case for short molten zone lifetimes.
• the rate of solidification is be ⁇ lieved to be a straightforward function so the error can be easily determined.
• redistribution will occur nearly proportionately.
• the timing circuit 24 integrates the time that the exposed region is molten and terminates the pulses when the prescribed period is complete. Alternatively the laser power can be adjusted. Such "integration" is necessary because, in some instances, discussed further below, resolidification of the molten region occurs after each pulse and prior to the application of the following pulse.
• the detector 22 is designed in a straightforward manner to convey the timing signal to timing circuit 24 only when the surface of the exposed region is reflective.
• the threshold for melting of the surface is first determined by exposing the surface to laser pulses of increasing energy until a change in surface reflectivity is measured, thus indicating the onset of melting. The details of the procedure are given above. The laser exposure is then continued at the threshold level or slightly above until the desired integrated exposure period is obtained, e.g., to obtain the desired annealing and/or the desired impurity profile. Depending on the laser pulse length a single pulse may be adequate, or many pulses may be used to give finer adjustment of the exposure time. This procedure is especially useful when the range between inadequate power to effect annealing, and excessive power that causes catastrophic damage, is narrow.
• Example _2 The size of the range, and its position in terms of power level, may vary significantly from sample to sample depending upon the depth and extent of damage, the surface characteristics, and the laser characteristics. It is largely this reason that makes it useful to monitor each sample, rather than exposing all samples with a similar processing history to a predetermined laser pulse.
• Example _2 The size of the range, and its position in terms of power level, may vary significantly from sample to sample depending upon the depth and extent of damage, the surface characteristics, and the laser characteristics. It is largely this reason that makes it useful to monitor each sample, rather than exposing all samples with a similar processing history to a predetermined laser pulse.
• Example _2 The size of the range, and its position in terms of power level, may vary significantly from sample to sample depending upon the depth and extent of damage, the surface characteristics, and the laser characteristics. It is largely this reason that makes it useful to monitor each sample, rather than exposing all samples with a similar processing history to a predetermined laser pulse.
• Example _2 The size of the range, and
• the duration of a single laser pulse which may alternatively be considered as a continuous laser beam, is controlled in response to changes in the surface reflectivity.
• the important measurement in this case is the onset of melting, again assumed to be variable from sample to sample.
• the amount of power necessary to initiate melting is a complex measure of the
• OMP power being absorbed by the sample which in turn is a function of several variables including the amount of damage, the concentration of impurities, the depth of both, and the surface characteristics of the sample. Therefore, a longer delay between the initial exposure to the laser and the onset of melting suggests the need for longer exposure to achieve annealing. It is, then, useful to measure the onset of melting and control the termination of the laser energy in accordance with the delay between the energization of the laser and that measurement. At this point we only know how to interrelate the delay and the necessary exposure time by empirical processes, as are straightforward in the art.
• Example 1 it was suggested that the use of many short pulses rather than one or a few pulses may give an added dimension to controlling the integrated exposure period.
• the pulse repetition rate would not matter since the procedure described here measures the integrated lifetime of the molten state at the surface of the sample and it would not matter whether the surface is continuously or intermittently molten.
• the surface characteristics change after melting and refreezing so that the power required to initiate melting after refreezing may be different from that needed initially.
• the damage anneals significantly during the first pulse which is likely if melting occurs, then the sample may not absorb sufficiently from subsequent pulses to remelt.
• This example involves an impurity redistribution during laser exposure.
• the sample initially is implanted with both boron, a p-type impurity, and arsenic, an n-type impurity.
• the substrate prior to implantation, has at least a surface layer relatively lightly doped n-type. Exposure of this structure to laser energy by the techniques described previously to give a molten state with a lifetime of 10 sec moves the boron deeper into the lightly doped n-layer than the slower moving arsenic, thus creating an n-p-n transistor structure with a base width of approximately 0.5 micrometers, and an emitter depth of approximately 1 micrometer.

Landscapes

• Recrystallisation Techniques (AREA)
• Physical Deposition Of Substances That Are Components Of Semiconductor Devices (AREA)
• Testing Or Measuring Of Semiconductors Or The Like (AREA)

Priority Applications (1)

Applications Claiming Priority (2)

Publications (1)

Family

ID=25467491

Family Applications (1)

Country Status (7)

Cited By (2)

Families Citing this family (34)

Citations (6)

• 1978
  • 1978-08-21 US US05/935,665 patent/US4155779A/en not_active Expired - Lifetime
• 1979
  • 1979-08-14 JP JP50164779A patent/JPS55500615A/ja active Pending
  • 1979-08-14 GB GB8009229A patent/GB2040077B/en not_active Expired
  • 1979-08-14 WO PCT/US1979/000834 patent/WO1980000509A1/en not_active Ceased
  • 1979-08-14 NL NL7920078A patent/NL7920078A/nl unknown
  • 1979-08-16 CA CA000333912A patent/CA1118114A/en not_active Expired
  • 1979-08-21 FR FR7921071A patent/FR2434483A1/fr active Granted

Patent Citations (6)

Non-Patent Citations (1)

Also Published As

Similar Documents

Legal Events