Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
MOLECULAR ION SOURCE FOR ION IMPLANTATION
Document Type and Number:
WIPO Patent Application WO/2013/068796
Kind Code:
A2
Abstract:
The present invention relates to the production of molecular phosphorous ion beams that can be used in low-energy ion implanters to manufacture semiconductor devices. More particularly, the present invention relates to the design and use of an ion source with a phosphine dissociator that can produce molecular P4 vapors from phosphine. The use of the gaseous phosphorous-hydrogen compound as a molecular P vapor source in combination with the dissociator speeds up the initiation and stabilization of injected gas into the discharge chamber from tens of minutes, in furnace (oven) fed systems, to seconds in the disclosed invention. The decomposition of phosphine into molecular phosphorous and hydrogen in the dissociator completely excludes the presence of P-H compounds at the output of the ion source, thus eliminating the problem of utilization of this toxic gas and simultaneously increasing the fraction of molecular P4 and P2 ions in the ion beam.

Inventors:
HERSHCOVITCH ADY (US)
GUSHENETS VASILIY (RU)
BUGAEV ALEXEY (RU)
KULEVOY TIMUR (RU)
OKS EFIM (RU)
Application Number:
PCT/IB2012/000830
Publication Date:
May 16, 2013
Filing Date:
March 16, 2012
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
BROOKHAVEN SCIENCE ASS LLC (US)
HERSHCOVITCH ADY (US)
GUSHENETS VASILIY (RU)
BUGAEV ALEXEY (RU)
KULEVOY TIMUR (RU)
OKS EFIM (RU)
International Classes:
H01J7/02
Attorney, Agent or Firm:
SACK, Alan et al. (3 World Financial CenterNew York, NY, US)
Download PDF:
Claims:
CLAIMS:

1. An ion source, comprising:

a dissociator that allows an input of a gas having a formula, XH3, where X is a group 15 element, wherein the dissociator splits the gas into a hydrogen and a molecular Xz gas, where z is 1 , 2, 3 or 4;

a discharge chamber connected to the dissociator comprising an emission opening, wherein the discharge chamber ionizes the molecular Xz gas; and an ion-optical system aligned with the emission opening of the discharge chamber that guides the ionized molecular Xz gas.

2. The ion source of claim 1, wherein X is P, As, or Sb.

3. The ion source of claim 1, wherein the gas is selected from the group

consisting of phosphine (PH3), arsine (AsH3), stibine (SbH3) and a combination thereof.

4. The ion source of claim 2, wherein the gas in the dissociator undergoes a reaction according to formula,

4XH3 = X4 + 6H2

5. The ion source of claim 4, wherein X is P.

6. The ion source of claim 4, wherein the X4 gas undergoes a dissociation

reaction in the discharge chamber according to formula,

X4 X4+ + e"

e" + X4+ 2e" + 2X2+

g° + X4+ -» e" + 2X2+

where g° is a UV photon.

7. The ion source of claim 4, wherein the hydrogen is dissipated through a

vacuum system.

8. The ion source of claim 1, wherein the dissociator comprises:

a heater spiral confined within a heat resistant tube; and a thermal shield surrounding the heat resistant tube with the heater spiral positioned within a holding chamber.

9. The ion source of claim 8, wherein the heat resistant tube and the heater spiral form an active zone of the dissociator.

10. The ion source of claim 8, wherein the heat resistant tube comprises quartz, aluminum or ceramic.

1 1. The ion source of claim 10, wherein the heat resistant tube comprises quartz.

12. The ion source of claim 1, wherein the dissociator is an RF dissociator.

13. The ion source of claim 1, wherein the discharge chamber comprises a

cathode, an anticathode, and an anode confined within a magnetic field, wherein the anode has a proximal end connected to the dissociator and a distal end comprising the emission opening.

14. The ion source of claim 13, wherein the anode is an extended hollow anode.

15. The ion source of claim 14, wherein the extended hollow anode has a shape of an extended rectangular parallelepiped.

16. The ion source of claim 1, wherein the emission opening is selected from the group consisting of a slit, an oval, a square, and a circle.

17. The ion source of claim 16, wherein the emission opening is a slit having dimensions of about 1 to 50 mm on the short side of the anode and about 2 to about 750 mm on the long side of the anode.

18. The ion source of claim 17, wherein the emission opening is a slit having dimensions of about 3 to 7 mm on the short side of the anode and about 40 to about 75 mm on the long side of the anode.

19. The ion source of claim 17, wherein the emission slit has dimension of about 1 x 40 mm2 with a longer dimension along the longest side of the anode.

20. The ion source of claim 13, wherein the cathode is a heated cathode.

21. The ion source of claim 20, wherein the heated cathode is an indirectly heated cathode.

22. The ion source of claim 20, wherein the heated cathode is a directly heated cathode.

23. The ion source of claim 22, wherein the directly heated cathode is a directly heated Penning-type cathode.

24. The ion source of claim 13, further comprising a shield near the cathode.

25. The ion source of claim 13, wherein the discharge chamber further comprises at least two diaphragm-type holes that provide access points for an arc discharge current generated by the cathode and the anticathode.

26. The ion source of claim 1, wherein the discharge chamber is a Bernas

discharge chamber, a Freeman discharge chamber, a Calutron discharge chamber, an Electron-Cyclotron Resonance (ECR) discharge chamber, or a microwave discharge chamber.

27. The ion source of claim 9, wherein the temperature of the active zone of the dissociator is not greater than about 1000 °C.

28. The ion source of claim 27, wherein the temperature of the active zone of the dissociator is between about 300 °C and about 1000 °C.

29. The ion source of claim 28, wherein the temperature of the active zone of the dissociator is between about 500 °C and about 900 °C.

30. The ion source of claim 29, wherein the temperature of the active zone of the dissociator is between about 700 °C and about 850 °C.

31. The ion source of claim 1, further comprising a vapor line connecting the dissociator and the discharge chamber.

32. The ion source of claim 1, wherein the vapor line comprises a heat resistant tube.

33. The ion source of claim 32, wherein the heat resistant tube is selected from the group consisting of a ceramic tube, a quartz tube and a stainless steel tube.

34. The ion source of claim 33, wherein the stainless steel tube of the vapor line has a diameter of about 5 mm and length of between 10 and 15 cm

35. The ion source of claim 1, wherein the ion-optical system comprises a

suppressing electrode and an accelerating electrode.

36. The ion source of claim 1, wherein the molecular Xz gas ions generated by the ion-source are used to implant a silicon wafer to produce an n-type semiconductor.

37. The ion source of claim 1, further comprising an additional gas or vapor source connected to the discharge chamber.

38. The ion source of claim 37, wherein the additional gas or vapor source is connected directly to the discharge chamber.

39. The ion source of claim 37, wherein the additional gas or vapor source is connected by a vapor line to the discharge chamber.

40. The ion source of claim 37, wherein the additional gas source is selected from N2, Ar, O2 or a combination thereof.

41. The ion source of claim 37, wherein the additional vapor source is selected from As, Sb, Ga, Ge or a combination thereof.

42. An ion source for generating molecular phosphorous ions from a phosphine gas, comprising:

a dissociator having a proximal end that allows an input of the phosphine gas and a distal end connected to a vapor line; wherein the dissociator splits the phosphine gas into a hydrogen and a molecular P4 gas;

a discharge chamber that comprises a heated cathode, an anticathode, and an extended hollow anode confined within a magnetic field, wherein the anode has a proximal end connected to the opposite end of the vapor line, and a distal end comprising an emission opening; wherein the discharge chamber ionizes the molecular P4 gas into P4+ and P2+ ions; and

an ion-optical system aligned with the emission opening near the distal end of the hollow anode that guides the molecular P4+ and P2+ ions.

43. The ion source of claim 42, wherein the dissociator comprises:

a heater spiral confined within a heat resistant tube; and a thermal shield surrounding the heat resistant tube with the heater spiral positioned within a holding chamber.

44. A method of producing a molecular ion beam comprising,

injecting a gas of phosphine (PH3), arsine (AsH3), stibine (SbH3) or a combination thereof into a dissociator;

splitting the gas in the dissociator into hydrogen and a molecular component selected from molecular phosphorus (P4), molecular arsenic (As4) or molecular antimony (Sb4);

directing the molecular component into a discharge chamber;

generating molecular ions from the molecular component in the discharge chamber by an arc discharge current; and

accelerating the molecular ions.

45. The method according to claim 44, further comprising directing the molecular ions to a bending magnet.

46. The method according to claim 44, wherein splitting of the gas in the

dissociator occurs according to a reaction,

4XH3 = X4 + 6H2

where X is P, As, or Sb.

47. The method according to claim 46, wherein X is P.

48. The method according to claim 46, wherein generating molecular ions from the molecular component in the discharge chamber occurs according to formula,

X4 X4+ + e"

e" + X4+ 2e~ + 2X2+

g° + X4+ -» e~ + 2X2+

where g° is a UV photon.

49. The method according to claim 44, wherein a temperature of the dissociator is not greater than about 1000 °C.

50. The method according to claim 44, further comprising directing an additional gas source selected from N2, Ar, 02 or a combination thereof into the discharge chamber;

generating a plurality of ions from the additional gas source in the discharge chamber by the arc discharge current; and

accelerating the additional gas ions.

51. The method according to claim 44, further comprising directing an additional vapor source selected from As, Sb, Ga, Ge or a combination thereof into the discharge chamber;

generating ions from the additional vapor source in the discharge chamber by the arc discharge current; and

accelerating the additional vapor ions.

52. A method of producing a molecular phosphorous ion beam comprising,

injecting a phosphine (PH3) gas into a dissociator;

splitting the phosphine (PH3) gas in the dissociator into hydrogen and a molecular phosphorous (P4);

directing the molecular phosphorus (P4) into a discharge chamber;

generating P4+ and/or P2+ ions from the molecular phosphorus (P4) in the discharge chamber by an arc discharge current; and

accelerating the P4+ and/or P2+ ions.

53. The method according to claim 52, further comprising directing the P4+ and/or P2+ ions to a bending magnet.

54. A method of ion implanting comprising

injecting phosphine (PH3) gas into a dissociator;

splitting phosphine into its hydrogen and molecular phosphorous components according to a reaction,

4PH3 = P4 + 6H2;

directing the molecular phosphorous component into a discharge chamber; generating molecular phosphorous ions in the discharge chamber by an arc discharge current;

accelerating the molecular phosphorus ions; directing the molecular phosphorous ions to a bending magnet to select ions of desired energy; and

impinging the molecular phosphorous ions of desired energy on a target.

55. The method of claim 54, wherein the target is a semiconductor.

56. The method of claim 55, wherein the semiconductor is a silicon wafer.

57. The method of claim 54, further comprising injecting one or more gas or vapor sources selected from (¾, N2, Ar, As, Sb, Ga, or Ge into the discharge chamber;

generating molecular or atomic ions from said gaseous or vapor sources in the discharge chamber by an arc discharge current;

accelerating the molecular or atomic ions from said gaseous or vapor sources; directing the molecular or atomic ions from said gaseous or vapor sources to a bending magnet to select ions of desired energy; and

impinging the molecular or atomic ions from said gaseous or vapor sources of desired energy on a target.

Description:
TITLE

Molecular Ion Source for Ion Implantation

CROSS-REFERENCE TO A RELATED APPLICATION

[0001] This application claims the benefit under 35 U.S.C. § 119(a)-(d) of a Russian

Patent Application Registration No. 201 1145645 filed on November 9, 2011, the content of which is incorporated herein in its entirety.

STATEMENT OF GOVERNMENT LICENSE RIGHTS

[0002] This invention was made with Government support under contract number

DE-AC02-98CH10886, awarded by the U.S. Department of Energy. The Government has certain rights in the invention.

BACKGROUND OF THE INVENTION

I. FIELD OF THE INVENTION

[0003] The invention relates to the generation of molecular ions for ion implantation in a semiconductor substrate. More particularly, the invention relates to the design and manufacture of an ion source with a gas dissociator that can convert phosphine gas into molecular phosphorous (P 2 & P 4 ) ions in safe and effective way. The invention also relates to the design and manufacture of an ion source with a gas dissociator that can convert arsine and stibine gas into molecular arsenic (As 2 & As 4 ) ions or molecular antimony (Sb 2 & Sb 4 ) ions.

II. BACKGROUND OF THE RELATED ART

[0004] The most basic building blocks of any semiconductor devices are n-type and p-type semiconductor devices. Such semiconductor devices include MOS, CMOS, MOSFET, diodes, transistors, integrated circuits, and the like. In manufacturing these devices, semiconductor substrates are implanted with dopants. Generally, the semiconductor substrate is made from a single crystal silicon wafer and the dopants may be atoms or molecules with properties that differ from those of the original semiconductor substrate. Once implanted, the dopants may alter the properties of the implanted regions such that the resulting substrate may have discrete regions with different properties. In addition to the implanted dopants, discrete regions with different properties in the substrate may be formed by specific ion implantation. For instance, the most widely used dopants for implantation are boron and phosphorous. Boron can be used to make the p-type semiconductors and phosphorous can be used to make the n-types semiconductors. After the boron and/or phosphorous dopants are implanted, other dopants can be implanted to form more complex devices.

[0005] One method of introducing dopants into a semiconductor substrate is through the use of an ion implanter which includes an ion source for converting a gas or a solid material into a well-defined ion beam. The ion beam typically is mass analyzed to eliminate undesired ion species, accelerated to a desired energy, and then the ions are implanted into a target. The ion beam may be distributed over the target area by electrostatic or magnetic beam scanning, by target movement, or by a combination of beam scanning and target movement. The ion beam is typically a spot beam or a ribbon beam. For speedy and efficient manufacturing of semiconductor devices, it is important to be able to switch quickly among various ion species for implantation. Therefore, pure gaseous sources are desired, since switching between species can be done in seconds while it takes close to an hour to achieve a stable vapor supply to the ion source and to heat and stabilize a furnace (oven) that produces the stable vapor supply.

[0006] One of such potential gaseous sources is phosphine gas that can be used as a working mixture in discharge systems of ion sources mainly to obtain both singly charged and multiply charged atomic P ions. The use of phosphine allows efficient control and stabilization of the gas flow rate to the discharge chamber. While, the use of phosphine (PH3) gas allows for efficient control and stabilization of the gas flow rate to the discharge chamber, the ion beam typically contains PH + and P¾ + ions along with P + , P 2+ , P 3+ , and P 4+ ions, and the amount of PH + and P¾ + ions can reach about 25-30 % of the total beam current. The presence of these ions in the ion beam, however, is undesirable because they decrease the gas efficiency of the ion source and increase the load on power supplies of its ion-optical system. Moreover, phosphorous-hydrogen ions are very toxic and corrosive, and therefore, ion implanters must be equipped with tools to utilize these ions present in the exhaust of vacuum pumps, otherwise the level of its discharge into the atmosphere can be intolerable.

[0007] With continued miniaturization of semiconductor devices, there has been an increased demand for decreasing the dimensions of the surface into which ions are implanted, thereby, increasing the density of elements per unit surface area of the substrate. However, increasing the density of elements per unit surface area has resulted in the need for shallow junctions. For example, decreasing the channel length and gate oxide thickness increases transconductance, the current drive of the transistor in metal oxide semiconductors. However, the decrease in the channel length also produces an undesirable short channel effect, which is a roll-off in threshold voltage. The short channel effect, however, can be minimized by decreasing gate oxide thickness and/or decreasing the junction depth.

[0008] Even though the channel lengths have been scaled aggressively over the last several years, the junction depth has not been scaled quite as aggressively, and in particular, ultra-shallow junctions have been hard to achieve with current technology. For example, decreasing junction depth can be achieved by decreasing ion energy. However, in all available ion implantation facilities, decreasing the ion energy, e.g., the accelerating voltage, automatically decreases the current of the transported ion beam due to the effect of its space- charge based on the Child-Langmuir relationship. Therefore, the retention of high ion current density responsible for the efficiency of the process is limited, resulting in low production rates. One example of the observed production rates are found in the implantation of single/atomic phosphorous ions generated from phosphine.

[0009] One of the possible ways to decrease the energy per dopant atom is to generate molecular ion beams that contain two or more dopant atoms, i.e., polyatomic conglomerates or clusters (M n ). In particular, the implanted energy per atom in a molecular ion at a given accelerating voltage decreases by a value multiple of the number of atoms (n) in the molecule or cluster. In contrast, the implanted dose increases by a square of the number of atoms (n 1'5 ) compared to the implanted energy of monatomic ions. Thus, the use of molecular ions instead of monoatomic ions is very promising for generating shallow junctions.

[0010] Horsky T. N. proposed generating molecular ions using furnace generated molecular phosphorous vapors from a solid red phosphorous and a conventional beam line implanter with a dual-mode ion source that operates in a cluster formation mode and a monomer formation mode. The cluster formation mode is mediated by electron impact ionization and the monomer formation mode is mediated by an arc discharge. ("Universal ion source for cluster and monomer implantation," AIP Conf. Proc, v. 866, 159 (2006); incorporated herein by reference in its entirety and referred to herein as "Horsky"). As illustrated in FIG. 1, the discharge chamber in the Horsky ion source is an extended rectangular parallelepiped with an ion extraction slit along its longest dimension. The electron gun creates an energetic electron beam, which is deflected by 90 degrees (90°) through the source by a magnetic dipole field. The deflected electron beam enters the source ionization chamber through a small entrance aperture. Once it is within the ionization chamber, the electron beam is guided along a path parallel to and directly behind the ion extraction slit by a uniform axial magnetic field produced by a permanent magnet surrounding the ionization chamber. Ions are thus created along the electron beam path and adjacent to the extraction slit. Implanter optics are located outside of the discharge chamber on the source side of the emission slit. According to Horsky, this serves to provide good extraction efficiency of the ions, such that an ion current density of up to 1 mA/cm 2 can be extracted from the source. The ion source system of Horsky, however, is difficult to operate due to its design complexity and the use of the electron gun with electron beam rotation.

[0011] To overcome this limitation, Gushenets et al. proposed generating molecular ions using furnace generated molecular phosphorous vapors from a solid red phosphorous and a hot-cathode ion source. ("Molecular Phosphorus Ion Source for Semiconductor Technology," Proc. of the 10th Int. Conf. on Modification of Materials with Particle Beams and Plasma Flows, Tomsk 2010, pp. 783; incorporated herein by reference in its entirety) As illustrated in FIG. 2, the Gushenets, et al. ion-source system has a discharge chamber with a hollow anode. The anode has an emission slit located along the side of the discharge chamber opposite the vapor entrance. Furthermore, the anode has two holes at the faces of the discharge chamber through which electrons emitted from the hot cathode and electrons are reflected from the anticathode. To produce the vapor source for the discharge chamber, Gushenets et al. describe a procedure of evaporating red phosphorus in a crucible with phosphorus in a furnace maintained at about 800 °C. Once the molecular phosphorous from the oven enters the discharge chamber, it comes in contact with a discharge plasma from the arc discharge current (i.e., an electron beam), which is guided along a path parallel to and directly behind the ion emission slit by a uniform axial magnetic field.

[0012] The ion source system of Gushenets et al. as well as Horsky, however, produce the phosphorous (P) ions, including molecular phosphorous ions, by an electron beam ionization of phosphorous vapors. These phosphorous vapors are supplied to the ionization chamber as a result of heating solid phosphorous in the furnace to produce a red phosphorous. Red phosphorous, however, has many undesirable qualities. For instance, red phosphorous is thermodynamically unstable. It is also capable of transforming from one allotropic state to another. Finally, red phosphorus has low kinetics of phosphorous evaporation and inconsistent evaporation rate. All of these factors lead to an unstable supply of phosphorous vapors in the discharge chamber of the ion source. Moreover, the low kinetics of phosphorous evaporation complicates the ability to control the vapor flow rate, which relies on varying the heating temperature of red phosphorous and can substantially lengthen its adjustment by the order of tens of minutes. For instance, the solid phosphorous has to be heated in the furnace gradually, otherwise a sublimation will result. Typically, it takes about 20 minutes to heat up and another 20 minutes to get gas discharge. Since it takes a substantial amount of time for the vapor pressure to reach equilibrium over the heated phosphorous surface, it becomes too difficult to obtain vapors of molecular phosphorous with a constant flow rate and, therefore, significantly complicates the automation of the system especially if multiple ions need to be implanted on the silicon wafer.

[0013] In view of the foregoing, it would be desirable to provide a solution which overcomes the above-described inadequacies and shortcomings in the production of molecular phosphorous ions for the low-energy ion implantation that can produce shallow junctions for semiconductor or other related industries without the need to use the furnace (oven).

SUMMARY

[0014] Having recognized the shortcomings of the prior art, as one embodiment, a molecular phosphorous (phosphorus) ion-source is provided. The ion source contains a dissociator that splits phosphine molecule(s) into hydrogen and molecular phosphorous components before they are delivered to a discharge chamber, preferably via a vapor line. In one particular embodiment, the dissociator has a heater spiral within a heat-resistant tube covered by a thermal shield. The dissociator converts the phosphine gas into the molecular phosphorous and advantageously provides on demand feed for the discharge chamber that permits the generation of phosphorous ions for low energy ion implantation. It also affords rapid switching (on the order of a few seconds) between species during the implantation. The use of the dissociator avoids production of a gaseous stream of molecular phosphorous by the undesirable process of heating the solid red phosphorous in a furnace (oven).

[0015] While the molecular phosphorous ions are generally used to produce n-type semiconductors, other atomic or molecular species in gaseous form can be used in the subsequent implantation to form more complex devices. For example, other ionic species can be derived from other gaseous sources such as (¾, N 2 , and Ar, or when gaseous species are not available, the ionic species can be derived from vapor sources such as As, Sb, Ga, and Ge vapors.

[0016] The disclosed discharge chamber is an extended hollow anode with an emission slit for ion extraction and two diaphragm-type holes at opposite ends. The diaphragm-type holes provide the access points for the arc discharge current generated by a directly heated Penning-type cathode at one end and an anticathode (electron reflector) at the opposite end. Once the molecular phosphorous from the phosphine dissociator enters the discharge chamber, it comes in contact with a discharge plasma from the arc discharge current, which is guided along a path parallel to and directly behind the ion emission slit by a uniform axial magnetic field produced by a permanent or an electro magnet surrounding the discharge chamber. The high operating temperature and/or the high discharge current density in the discharge chamber causes the molecular phosphorous to ionize. The ion-source system further has a suppressor with an accelerating electrode located outside of the discharge chamber at opposite ends of the emission slit. In low-energy ion implantation, after accelerating the molecular phosphorous ions by the disclosed ion source, the molecular phosphorous ions are preferably directed into a bending magnet to select desired ions of a desired energy, which are then used to impinge the silicon wafer target.

[0017] In another embodiment, a method of producing a molecular phosphorous ion beam is provided. The method employs the steps of injecting phosphine (P¾) gas into a dissociator; splitting phosphine into hydrogen and molecular phosphorous components free of Pt¾ and PH; directing the phosphorous component generated in the dissociator into a discharge chamber, preferably via a short vapor line; generating molecular phosphorous ions in the discharge unit by a collision of electrons with molecular phosphorous; and collecting the generated molecular phosphorous ions. In yet another embodiment, a method of low- energy ion implantation is provided based on the production of molecular phosphorous ions.

[0018] The use of the gaseous phosphorus as a molecular P vapor source in combination with the dissociator makes it possible to speed up the adjustment of the flow rate from tens of minutes to a few seconds and to stabilize the gas supply. The decomposition of phosphine into P 4 and molecular hydrogen in the dissociator makes it possible to completely exclude the presence of phosphorous-hydrogen compounds at the output of the ion source. This fully eliminates the problem of utilization of the toxic PH + and Pt¾ + ions and simultaneously increases the fraction of molecular P 4 and P 2 ions generated by the ion source.

[0019] The objectives, features and advantages of the disclosed invention will be apparent from the following detailed description, which is to be read in conjunction with the accompanying drawings. The scope of the invention will be pointed out in the claims. The following drawings, taken in conjunction with the subsequent description, are presented to enable one of ordinary skill in the art to make and use the invention and to incorporate it in the context of particular applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] FIG. 1 schematically illustrates the ion source proposed by Horsky (2006).

[0021] FIG. 2 schematically illustrates the ion source proposed by Gushenets et al. (2010).

[0022] FIG. 3 schematically illustrates an exemplary gas-discharge molecular ion source with a phosphine dissociator.

[0023] FIG. 4 illustrates one embodiment of a phosphine dissociator.

[0024] FIG. 5 is an ion beam spectrum showing ion beam composition with phosphine flow under three different temperatures of the heater. The insert window is an enlarged view of the first major peak that shows the influence of temperature of dissociator on phosphine decomposition.

DETAILED DESCRIPTION

[0025] A gas-discharge molecular phosphorous ion source is provided that can be used in low-energy ion implanters to manufacture shallow junctions in semiconductor devices including highly integrated circuits. The present ion source can safely and effectively produce molecular phosphorous ion beam(s) from phosphine (PH3) gas without the need to create a gaseous ion stream by heating solid phosphorous source in the oven. In general, the ion source has a discharge chamber and a phosphine dissociator interconnected by a vapor line. The dissociator splits phosphine (PH 3 ) molecule(s) into their hydrogen and phosphorous components according to the reaction (1).

4PH 3 P 4 + 6H 2 (l) These product compounds enter the discharge chamber, and once in the discharge chamber, it is believed that P 4 molecules undergo a dissociation reactions (2)-(4):

P 4 ÷ P 4 + + e " (2) e ~ + P 4 + -_> 2e ~ + 2P 2 + (3) g° + P 4 + ÷ e + 2P 2 + (4) where g° is a UV photon. The hydrogen gas does not dissociate and typically is dissipated through a vacuum system.

[0026] An exemplary ion source is shown in FIG. 3. It has a number of individual components or subsystems: (1) Ρ¾ gas as a source of molecular phosphorous; (2) dissociator 10; (3) discharge chamber 20; and (4) vapor line 30. Each component or subsystem will be considered and described in detail herein below. In one embodiment, the ion source is configured as a provider for low-energy ion implantation in highly integrated circuits, such as semiconductors. For instance, the molecular phosphorous ions can be used to produce n-type semiconductors from a solid silicon wafer, while other atomic or molecular species in gaseous form can be used in the subsequent implantation to form more complex devices.

[0027] It is also contemplated that instead of using phosphine gas source, other trihydride group 15 gases can be used due to their molecular similarity to phosphine. For example, an arsine (Astt) gas or a stibine (Sbtt) gas can be used because in a similar fashion to phosphine, it is contemplated that the dissociator can split arsine (Astt) and stibine (Sbtt) molecule(s) into their hydrogen and arsenic or antimony components according to the reaction (5) and (6), respectively.

4AsH 3 As 4 + 6H 2 (5) 4SbH 3 Sb 4 + 6H 2 (6) These product compounds enter the discharge chamber, and once in the discharge chamber, it is believed that As 4 and/or Sb 4 molecules similarly undergo a dissociation reactions (2)-(4). The hydrogen gas does not dissociate and typically is dissipated through a vacuum system. A source of molecular phosphorous

[0028] The source of phosphorus, including molecular phosphorus, used in generation of phosphorous clusters and molecular ions for low-energy ion implantation is preferably phosphine. Phosphine (IUPAC name: phosphane) is a gaseous compound of phosphorus and hydrogen under normal conditions having a chemical formula PH 3 . It is a colorless, flammable, toxic gas and if pure it is also odorless.

[0029] Phosphine may be prepared in a variety of ways (A. D. F. Toy, The Chemistry of Phosphorus, Pergamon Press, Oxford, UK, 1973; incorporated herein by reference in its entirety) or purchased in pure form (electronic grade; >99.9995%), for example, from Sigma- Aldrich, St. Louis, MO. Industrially it can be made by the reaction of white phosphorus with sodium hydroxide, producing sodium hypophosphite and sodium phosphite, as a by-product. Alternatively the acid-catalyzed disproportioning of white phosphorus may be used, which yields phosphoric acid and phosphine. Both routes have industrial significance. However, the acid route requires additional purification and pressurization steps. Although, the acid route is a preferred method in cases when further reaction of the phosphine to substituted phosphines is needed. Phosphine can also be made by the hydrolysis of a metal phosphide, such as aluminum phosphide or calcium phosphide, or a reaction of potassium hydroxide with phosphonium iodide (PH4I), which can generate pure phosphine, free from P2H4. Since technical grade phosphine has traces of P2H4, and the mixture is spontaneously flammable in air, burning with a luminous flame, it is preferred that the supplied phosphine is pure and free of P2H4. However, it is also envisioned that phosphine with trace amounts of other materials may be used under oxygen free environment to avoid spontaneous combustion. [0030] As illustrated in FIGs. 3 and 4, the phosphine gas in one exemplary embodiment is injected into a dissociator 10 through a specifically designed valve. In one embodiment, the amount of phosphine supplied to a dissociator is between about 1 to about 10 seem (standard cubic centimeter per minute). In a preferred embodiment, the amount of phosphine supplied to a dissociator is about 3 to about 6 seem.

Dissociator

[0031] As illustrated in FIG 3 the phosphine gas initially enters the phosphine dissociator 10 where the phosphine is split into its phosphorous and hydrogen components according to reaction (1). Although the dissociator can have other components and elements, in general, the dissociator 10 has a chamber 14, a thermal shield 13, a heat resistant tube 12, and heater spiral 11. The space between the heat resistant tube 12 and the heater spiral 11 is considered an active zone of the dissociator 10. Preferably, the active zone of the dissociator 10 is heated to no greater than 1000 °C, suitably between 300 °C and 1000 °C, preferably between 500 °C and 900 °C, and most preferably between 700 °C and 850 °C.

[0032] One exemplary embodiment of the dissociator 10 is illustrated in FIG. 4. In this embodiment, the dissociator 10 has a heater spiral 11 confined within a heat resistant tube 12. The heater spiral 11 can be made from various materials, such as tungsten, iron- chrome aluminum, nickel-chrome, nickel-iron, nickel, stainless steel, copper, molybdenum, and M0S1 2 / MoSi. Preferably, the heater spiral 11 is made from a tungsten wire having a diameter of about 0.4 mm. The maximum value of the heated current supplied to the heater spiral 11 can be as high as 2.4 A, although the current is preferably set to a value between 1.8 A to 2.4 A. The heat resistant tube 12 can be made from any heat resistant material such as quartz, aluminum, and ceramic. In a preferred embodiment, the heat resistant tube 12 is made from high quality quartz. The heater spiral 11 and the heat resistant tube 12 form an active zone of the dissociator 10 in which phosphine flows and dissociates. The length of the active zone is defined as the length of the heater spiral 11 confined by the heat resistant tube 12 and is preferably between 10 cm and 30 cm, and more preferably about 20 cm. In one exemplary embodiment, the length of the active zone is about 19 cm. The heat resistant tube 12 with the heater spiral 11 is placed within a chamber (14 in FIG. 3) having a cylindrical tube 14a made of resilient material, such as stainless steel, running along the longitudinal axis of the heat resistant tube 12. The rims of the tube 14a at opposite ends support two flanges 14b and 14c to form the body of the chamber identified as 14 in FIG. 3. The tube 14a can have a diameter of about 10 mm to 100 mm, although, 30 mm diameter is more preferred in this particular embodiment. The heat resistant tube 12 is attached to a support element 16 that sits on top of the flange 14b to avoid any contact between the cylindrical tube 14a of the chamber and the heat resistant tube 12. The heat-resistant tube 12 is also preferably thermally insulated from the tube 14a by thermal screen 13, which can be a thin titanium foil with a mirror polished surface. However, to avoid overheating of the system, the cylindrical tube 14a is connected to an arc cooling radiator 15. On one side of the tube 14a is an active zone housing 17, which is connected to a vapor drift line 30 (see FIG. 3) through a metal- ceramic valve 19. The other the tube 14a has a gas feed cover 18 attached to a gas flow valve, e.g., "Micromate" made by Hoke Inc. (Spartanburg, SC) that supplies the phosphine gas. Thus, once the phosphine gas enters the dissociator 10, it is split into molecular hydrogen and phosphorus in the form of vapor consisting almost entirely of P 4 molecules that are further delivered via vapor line 30 to the discharge chamber 20. Although, the parameters of the dissociator are discussed with reference to the discharge chamber illustrated in FIG. 3, it is understood by those skilled in the art that the size, shape, form, etc. of the dissociator depends on the overall system and can be adjusted accordingly. For instance, if the size of the discharge chamber is modified, it is also contemplated that the size of the dissociator may have to be adjusted as well. For example, if the discharge chamber is increased in size (dimension) to accommodate the need for more molecular ions, the dissociator can be proportionally increased to accommodate the prescribed discharge chamber.

[0033] It is also contemplated that other dissociators, such as an RF dissociator described in Hershcovitch et al. (Rev. of Set Instrum. 58, 547, 1987; the disclosure of which incorporated herein by reference) and Hershcovitch (Phys. Rev. Lett. 63, 750 (1989); the disclosure of which incorporated herein by reference) may also be used in the disclosed ion source to generate molecular phosphorous gaseous stream.

Discharge chamber

[0034] Once the phosphine gas is split into its phosphorous and hydrogen components, the two gases are allowed to enter the discharge chamber. As illustrated in FIG. 3, the ion-source system utilizes a similar discharge chamber proposed by Gushenets et al. and illustrated in FIG. 2. ("Molecular Phosphorus Ion Source for Semiconductor Technology," Proc. of the 10th Int. Conf. on Modification of Materials with Particle Beams and Plasma Flows, Tomsk 2010, pp. 783; incorporated herein by reference in its entirety). In particular, as shown in FIG. 3, the discharge chamber 20 has a hollow anode 21 shaped as an extended rectangular parallelepiped. The dimensions of the anode 21 can vary widely depending on desired implanter ion beam cross section.

[0035] In one exemplary embodiment, the dimensions of the anode 21 are about 8 mm x 14 mm x 70 mm. The anode 21 has an emission slit 28 located along the side of the discharge chamber 20 opposite the vapor entrance, which is connected to the vapor line 30. Preferably, the slit 28 covers at least 50 % of the anode 21 in length. In one embodiment, depending on the overall length of the anode 21, the length of the slit 28 can range between about 2 mm and about 750 mm, or between about 40 mm and about 75 mm. In one exemplary embodiment, the length of the slit 28 is about 40 mm. While the slit 28 can be long as compared to the length of the anode 21, the slit 28 is preferably substantially narrow as compared to the depth of the anode 21. In one embodiment, the width of the slit can range between about 1 mm and about 50 mm or between about 3 and about 7 mm. In one exemplary embodiment, the slit 28 is about 1 mm wide. Furthermore, the anode 21 has at least two diaphragm-type holes (unmarked), one each at the opposite ends of the anode. One hole is near a heated Penning-type U-shaped filamentary cathode 22 and a shield 23, and the other hole is near an electron reflector or anticathode 24. The diaphragm-type holes are preferably about 3 x 6 mm 2 and provide the access points for the arc discharge current generated by the cathode 22.

[0036] Once the molecular phosphorous from the phosphine dissociator 10 enters the discharge chamber 20, it comes in contact with a discharge plasma from the arc discharge current, which is guided along a path parallel to and directly behind the ion emission slit 28 by a uniform axial magnetic field produced by a permanent or an electro magnet 25 surrounding the discharge chamber 20. The high operating temperature and/or the high discharge current density in the discharge chamber 20 causes the molecular phosphorous to dissociate into molecular phosphorous ions. During production of molecular phosphorous ions, the anticathode 24 is connected to the anode 21. The ion-source system further has an ion-optical system located outside of the discharge chamber at opposite ends of the emission slit 28.

[0037] The ion-optical system is a standard two-electrode setup known in the art. In particular, it has a suppressor (deceleration) electrode 26 and accelerating electrode 27 that are aligned with the emission slit 28. In low-energy ion implantation, after accelerating the molecular phosphorous ions by the present ion source, the molecular phosphorous ions in the ion beam 40 are preferably directed into a bending magnet (not shown) to select ions of a desired energy, which are then used to impinge the target, such as the silicon wafer. In one embodiment, the ion beam 40 can be accelerated to a voltage of between about 400 V and about 30 kV, depending on the desired depth of implantation.

[0038] Alternatively, instead of employing the discharge chamber 20 to generate the molecular ion beam, other discharge chambers are also contemplated for production of an ion beam for semiconductor implantation such as Bernas discharge chamber, Freeman discharge chamber, Calutron discharge chamber, Electron-Cyclotron Resonance (ECR) discharge chamber, and microwave discharge chamber. These discharge chambers are thoroughly described in "Atomic, Molecular, and Optical Physics: Charged Particles" ed. Dunning and Hulet, Vol. 29A, Academic Press, California, 1995 (ISBN 0124759742) pages 69-168, Matsuo, et al., AIP Conference Proceedings 1321, proceedings of the 18 th Ion Implantation Technology 2010 Kyoto Japan 6-11 June 2010, and Hershcovitch et al., Rev. Set Instrum. 77, 03B510 (2006), the disclosure of which incorporated herein by reference.

Vapor line

[0039] As illustrated in FIG. 3, the vapor line 30 is a tube made from a resilient material, such as stainless steel. The length of the vapor line 30 is not particularly limited but preferably is chosen to be as small as possible permitted by the ion source design. This is necessary to keep the walls of the vapor line 30 at a sufficiently high temperature, due to the heat released in the discharge chamber 20 and active zone of the dissociator 10, to preclude phosphorus vapor condensation on the walls of the vapor line 30. At the same time, the length of the vapor line 30 cannot be too small because the vapor line 30 ensures a required pressure difference between the active zone of the dissociator 10 and the discharge chamber 20. In one exemplary embodiment, the length of the vapor line with diameter of about 5 mm is between 10 cm to 15 cm, depending on the implanter selected and the number of other gas sources that are added to the vapor line. In addition to the vapor line 30 shown in FIG. 3, other vapor lines (not shown) can be attached to directly to the discharge chamber 20 or through a valve 31 to provide a supply of other atomic or molecular species in gaseous form to be used in the subsequent implantation. For example, the molecular phosphorous can be used to prepare n-type semiconductors, while other ionic species derived from gaseous sources such as (¾, N 2 , and Ar, or vapor sources such as As, Sb, Ga, and Ge, can be used to form more complex devices.

Method of producing molecular phosphorus for low-energy implanters

[0040] In one embodiment, a method of producing molecular phosphorus ion beams that can be used in low-energy ion implanters to manufacture highly integrated circuits is provided. The method includes the steps of injecting phosphine gas into a dissociator; splitting phosphine (PH 3 ) molecule(s) into its hydrogen and phosphorous components; directing the phosphorous components into the discharge chamber; generating P 4 + and/or P2 + ions in the discharge unit by an arc discharge current; accelerating P 4 + and/or P2 + ions; and directing P 4 + and/or P2 + ions to a bending magnet.

[0041] In yet another embodiment, a method of ion implantation using the molecular phosphorous ions is provided. An ion implantation system typically includes an ion source to produce the molecular or monoatomic ions, an accelerator to increase the energy of the generated ions, a separation magnet to select the ion(s) of a desired energy, and a target chamber, where the generated ions impinge on a target to produce, for example, shallow junctions in silicon wafers. Initially, molecular P ions are generated and accelerated in the ion source illustrated in FIG 3. The generated ions are then passed through a bending (separation) magnet well known in the art to select ions of only desired mass energy profile, and then they are allowed to impact the preselected target, e.g., silicon wafer to convert the silicon wafer into an n-type semiconductor. EXAMPLES

Example 1

[0042] An ion-source system was designed and tested at the High Current Electronics

Institute in Tomsk, Russia. A schematic illustration of the designed ion source system is shown in FIG. 3. The production of molecular phosphorous from phosphine was examined using the ion-source system. The dissociator was designed according to FIG. 4. It had a stainless steel tube of 30 mm diameter and two flanges crossed the two end of the tube. One end of the tube had a gas flow "Micromate" valve made by Hoke Inc. (Spartanburg, SC). The other end was connecter to vapor drift line, which provided gas flow to the discharge unit. Inside the stainless steel tube there was quartz tube with a heater spiral. The quartz tube was thermally insulated from the stainless steel tube by thermal screen made from thin titanium foil with mirror polished surface. The heater spiral was made from a thin tungsten wire of 0.4 mm in diameter. The heater spiral and the quartz tube formed an active zone of a dissociator in which phosphine flows. The length of the active zone was 19 cm.

[0043] The discharge chamber was designed according to Gushenets et al. (2010). In particular, the discharge chamber has a hollow anode shaped as an extended rectangular parallelepiped with dimensions of 8 mm x 14 mm x 70 mm. The anode has an emission slit located along the side of the discharge chamber opposite the vapor entrance, which is connected to the vapor line. The emission slit was set to 1 x 40 mm 2 . Furthermore, the anode has at least two diaphragm-type holes at the opposite ends, one near a heated Penning-type U- shaped filamentary cathode provided with a shield, and the other near an electron reflector or anticathode. The diaphragm-type holes were set to 3 x 6 mm 2 . The ion source also has a permanent or an electro magnet surrounding the discharge chamber that produces a uniform axial magnetic field. During production of molecular phosphorous ions, the anticathode was connected to the anode. The ion-source system further has an ion-optical system located outside of the discharge chamber at opposite ends of the emission slit. It has a suppressor (deceleration) electrode and accelerating electrode that are aligned with the emission slit.

Example 2

[0044] The ion source was tested under 300 V discharge burning voltage with the discharge current set to about 120-130 mA. An ion extracting beam current was set to 0.6 mA. In the ion-optical system , the ion accelerating voltage for the electrode 27 was set to 15 kV and the suppression voltage for the electrode 26 was set to -4.5 kV. The gas flow rate was not measured but the working gas pressure was monitored in vacuum vessel near the entrance of a bending magnet. The working gas pressure was as low as 6 x 10 "5 Torr. These parameters correspond to the parameters employed by Gushenets et al. (2010) for the optimal operation of the ion source carried out with using an oven based red phosphorous vapor source.

[0045] Maximum value of the heated current in the dissociator was set to 2.4 A. The current voltage drop on the spiral heater reached about 20 V that is equivalent to the temperature of the heater of about 820 °C.

Example 3

[0046] The ion source operated in a so called "beam plasma" mode when the anticathode electrode 4 was electrically connected to the anode 1 of discharge system (as illustrated in FIG. 3) in order to test splitting of molecule of phosphine (Pt¼) into its hydrogen and phosphorous components.

[0047] FIG. 5 shows an ion beam spectrum. In particular, FIG. 5 shows the ion beam composition for three different temperatures of the dissociator. The temperature was estimated based on the current supplied to the heater spiral. Measurements of the ion composition were made, step-by-step, after each increase in the current. The line 1 in FIG. 5 illustrates a case where the heater spiral current is relatively low, i.e., less than 1.6 A, and process of phosphine dissociation is not optimal. The line 2 illustrates the situation where the heater spiral current is about 1.7 A. The line 3 illustrates the situation where the heater spiral current is about 1.8 A. When the heater spiral current was set above 1.8 A further changes in the ion beam composition were not observed.

[0048] Since there is small remnant magnetization of the bending magnet, hydrogen was not observed even though the extraction energy was increased to 30 kV. It is believed that hydrogen was not ionized, since ionization potential of molecular and atomic hydrogen are 15.37 eV & 13.58 eV respectively according to W. Bleakney PR 40, 496 (1932). In contrast, the ionization potential for phosphine is only 10.2 eV (Y. Wada et al Inorganic Chemistry 3, 174 (1964)), while the ionization potential for P 4 is 9.2 eV (C. Brundle et al Inorganic Chemistry 1 1, 20 (1972)). Thus, it is believed that the hydrogen is not ionized and is simply pumped out by the implanter vacuum pumps. The avoidance of hydrogen ionization can adventurously provide a reduction in the power requirements and a heat load associated with the extraction and beam optics.

[0049] An insert in FIG. 5 shows that an increase in the dissociator temperature causes a nearly full disappearance of PH + and PI¾ + fractions in the generated ion beam. This demonstrates a high efficiency in the operation of the dissociator. In particular, the use of the dissociator in the ion source ensures almost doubling of P2 and P 4 ions and almost complete absence of phosphorous-hydrogen compounds in the ion beam. The absence of phosphorous-hydrogen compounds in the ion beam solves the problem of utilization of phosphine because these compounds are potential environmental hazards and threat to attending personnel. Moreover, the use of the gaseous compound of phosphorus with hydrogen as a molecular phosphorus vapor source in combination with the dissociator makes it possible to stabilize the gas supply and to speed up the adjustment of the flow rate from minutes to seconds.

Example 4

[0050] The ion source system was compared to Clusterlon® source shown in FIG. 1, which was developed by Horsky (2006). The Clusterlon® source has an oven for generation of the molecular phosphorous vapor from the red phosphorous and a discharge chamber to generate molecular phosphorous ions. The discharge chamber of Horsky has an anode with an emission slit having dimensions of 8 x 48 mm that can provide a phosphorous ion current close to 3 mA. The emission surface area in the Clusterlon® Source is an order of magnitude larger than the emission surface area in the ion source described in Example 1 , however, both sources have nearly the same emission current density of about 1 mA/cm 2 . Thus the plasma parameters inside the discharge chambers for Clusterlon® source and the ion source described in Example 1 are identical. However, the ion source described in Example 1 and illustrated in FIG. 3 has a much simpler design because it does not use an electron gun with a bending magnet. Moreover, the ion source described in Example 1 and illustrated in FIG. 3 almost completely eliminates PH and P¾ ions in the ion beam. Thus, the present ion source increases the gas and energy efficiency and fully eliminates the problem of phosphine utilization.

[0051] It will be appreciated by persons skilled in the art that the present invention is not limited to what has been particularly shown and described. Rather, the scope of the present invention is defined by the claims which follow. It should further be understood that the above description is only representative of illustrative examples of embodiments. The description has not attempted to exhaustively enumerate all possible variations. The alternate embodiments may not have been presented for a specific portion of the invention, and may result from a different combination of described portions, or that other undescribed alternate embodiments may be available for a portion, is not to be considered a disclaimer of those alternate embodiments. It will be appreciated that many of those undescribed embodiments are within the literal scope of the following claims, and others are equivalent. Furthermore, all references, publications, U.S. Patents, and U.S. Patent Application Publications cited throughout this specification are hereby incorporated by reference as if fully set forth in this specification.