CA1340058C - Magnetooptical recording medium allowing overwriting with tow or more magnetic layers and recording method utilizing the same - Google Patents

Abstract

A magnetooptical recording medium provided with a first magnetic layer and a second magnetic layer having a higher Curie point and a lower coercive force than those of said first magnetic layer and being exchange coupled with said first magnetic layer is disclosed. That satisfies a relation:

(see fig. I) wherein H H is the coercive force of the first magnetic layer; H L is the coercive force of the second magnetic layer; M S is the saturation magnetization of the second magnetic layer; h is the thickness thereof;
and .sigma. w is the magnetic wall energy between the first and second magnetic layers.

Description

n~

~ TITLE OF THE INVENTION
Magnetooptical Recording Medium Allowing Overwriting With Two or More Magnetic Layers and Recording Method Utilizing the Same BACKGROUND OF THE INVENTION
Field of the Invention The present invention relates to a magneto-optical recording medium provided with a recording layer composed of a magnetic film allowing information record-ing by irradiation with a light beam, and a recording method utilizing said recording medium.
Description of the Related Background Art Optical memory devices utilizing laser beam are being actively developed in recent years as memories of high density and large capacity. Particularly magnetooptical recording is attracting attention as a rewritable recording method, and the magnetooptical recording medium employed in such recording method is greatly anticipated as a rewritable optical memory device.
Fig. 1 schematically illustrates a conven-tional apparatus for such magnetooptical recording, wherein a disk-shaped magnetooptical recording medium or magnetooptical disk 31, provided with a magnetic layer having an easy axis of magnetization perpendicular to the surface of said layer, is rotated by a spindle 13~100:~

motor 32. An optical head 34, provided with a laser unit, an objective lens 34 etc., performs information recording by projecting a light beam 35, turned on and off according to the information to be recorded, onto the disk 31 through the objective lens 33. A bias magnetic field is applied by an electromagnet 36 to an area of the disk 31 irradiated by the light beam.
The optical head 34 is moved by an unrepresented mechanism in the radial direction of the disk 31, thereby recording information in spiral or concentric patterns.
In conventional apparatus as shown in Fig. 1, the information recording and erasure are conducted in steps 30a to 30f shown in Fig. 2. At first, as shown in 30a, the magnetic layer 37, constituting the record-ing layer of the magnetooptical disk, is magnetized in a predetermined direction. Then, as shown in 36b, the magnetic layer 37 is irradiated by the light beam 35.
The irradiated area is heated close to the Curie point of said magnetic layer 37 by absorption of the irradi-ating beam, thus causing a decrease in the coercive force. In this state the magnet 36 shown in Fig. 1 applies a bias magnetic field B' of a direction opposite to the aforementioned predetermined direction, whereby the magnetization in the area irradiated by the light beam is inverted. Thus, after having passed the position of irradiation, there is formed, ~. , , .,. . , .. " ,. ,~ , , ,, ~,.

13'100~

1 as shown by 30c, a record bit 38 having a direction of magnetization different from that in the surrounding area. The information is recorded as a train of such record bits 38 or an information track.
Then, for erasing the information recorded as in 30d, an unmodulated light beam 35 is projected while a bias magnetic field -B' of a direction opposite to that of the magnetic field at the recording is applied by the magnet 36 as shown in 30e, thereby heating the magnetic layer 37 again to a temperature close to the Curie point. Thus the magnetic layer 37 restores the magnetization aligned in the predeter-mined direction, thus returning to the state prior to the recording as shown in 30f.

Also the recorded information can be reproduced by irradiating the magnetic layer 37 having record bits 38, with an unmodulated light beam of a reduced inten-sity insufficient for heating to the Curie point, and detecting the direction of polarization of the reflected or transmitted light beam by a known method utilizing magnetooptical effect.
However, in case of rewriting already recorded information, the conventional apparatus as explained above is incapable of so-called overwriting but requires a step of erasing followed by a step of new recording. Thus, in case of changing the infor-mation recorded in a track on a magnetooptical disk, " , . ~ . . ,~ . ". ", " . , 1~-100~

1 it becomes necessary to erase the information of said track in a turn of the disk and to record the new information in a succeeding turn, and such operation inevitably results in a loss in the recording speed.
In order to resolve such drawback there has already been proposed an apparatus equipped with a record/reproducing head and a separate erasing head, or an apparatus in which the recording is achieved by modulating the applied magnetic field while a continuous laser beam is projected, but such apparatus is associated with other drawbacks such as being bulky and expensive or being incapable of high speed modula-tion.
On the other hand, in order to improve the recording sensitivity and the S/N ratio at reproduction in such magnetooptical recording medium, a technology utilizing mutually exchange-coupled two magnetic layers is disclosed in the Japanese Patent laid-open ~1q~ //4 No. 78652/1982, corresponding to the U.S. Patent ~ li~ation Sor. ~'o. 315,~67 ~hieh i~ e~.~tinucd ~5 Co~ u~Liul~-in part No. 6~ 3, which is fu. Lhcl co~Lif,~lcd a~
~~nr.ti,lu~tio" N~. 90n,9~4. In addition to the above-mentioned patent ar~lication, such magnetic layer of two-layered structure was described in "Magnetization Process of Exchange-coupled Ferrimagnetic Double-Layered Films", Kobayashi et al., Japanese Journal ûf Applied Physics, Vol. 20. No. 11, November 1981, 5 ~ 13~00-~

l P. 2089 - 2095 and "Thermomagnetic Writing on Exchange-coupled Amorphous Rare-Earth Iron Double-layer Films"
Tsunashima et al., IEEE Transactions on Magnetics, Vol. MAG-17, No. 6, November 1981, P. 2940 - 2842.
However such exchange-coupled double-layered films are still incapable of overwriting and require an erasing step.

SUMMARY OF THE INVENTION
The object of the present invention is to provide a magnetooptical recording medium and a record-ing method, which are free from the above-explained drawbacks and enable an overwriting operation as in the magnetic recording media, by merely attaching magnetic field generating means of a simple structure to the conventional apparatus.
The foregoing object can be achieved, according to the present invention, by a magneto-optical recording medium composed of a substrate, a first magnetic layer formed thereon, and a second magnetic layer exchange-coupled with said first magnetic layer and having a higher Curie point and a lower coercive force compared with those of said first magnetic layer, and satisfying a condition:
HH ~ H ~ ~ /2M h wherein Ms is the saturation magnetization of the - 6 - 1~ 4 005~

1 second magnetic layer, h is the thickness thereof, ~w is the magnetic wall energy between two magnetic layers, and HH and HL are coercive forces the first and second magnetic layers.
The information recording on said medium is conducted by a step of applying a first magnetic field of a magnitude enough for magnetizing said second magnetic layer but insufficient for inverting the direction of magnetization of said first magnetic layer, and a step of applying a bias magnetic field of a direction opposite to that of said first magnetic field and simultaneously projecting a light beam of a power enough for heating the medium close to the Curie point of the first magnetic layer thereby obtaining the recording of a first kind in which the magnetization of the first magnetic layer is aligned in a direction stable to the second magnetic layer while the direction of magnetization of the second magnetic layer is not changed, or applying said bias magnetic field and simultaneously project-ing a light beam of a power enough for heating the medium close to the Curie point of the second magnetic layer thereby obtaining the recording of a second kind in which the direction of magnetization of the second magnetic layer is inverted and the first magnetic layer is simultaneously magnetized in a direction stable to the second magnetic layer.

13~0.j8 Fig. l is a schematic view of a conventional magnetooptical recording apparatus;
Fig. 2 is a schematic view showing the recording and erasing processes utilizing a conven-tional magnetooptical recording medium;
Fig. 3 is a schematic cross-sectional view of a magnetooptical recording medium embodying the present invention;

Fig. 4 is a chart showing the temperature characteristic of the coercive force of the magnetic layer in the medium shown in Fig. 3;
Fig. 5 is a schematic view showing the state of magnetization in a recording process utilizing the medium shown in Fig. 3;
Fig. 6 is a schematic view of an apparatus for recording and reproduction with the medium of the present invention;
Figs. 7A to 7D are charts showing B-H loops of the medium shown in Fig. 3;
Fig. 8 is a schematic view of an embodiment of the magnetooptical recording apparatus utilizing the medium of the present invention;
Fig. 9 is a chart showing the mode of modulation of the light beam irradiating the medium in the apparatus shown in Fig. 8;
Fig. l0A is a schematic view showing states - 8 - 13400~

1 of magnetization in the recording process utilizing the apparatus shown in Fig. 8;
Fig. lOB is a schematic view showing states of magnetization in the recording process utilizing another embodiment of the medium;
Fig. 11 is a chart showing the relation between the coercive force of the first magnetic layer and the magnetic field inducing noise increase;
Figs. 12 and 13 are schematic cross-sectional views of embodiments of the magnetooptical recording medium provided with an adjusting layer for the magnetic wall energy;
Fig. 14 is a chart showing the relation between the thickness of the adjusting layer and the exchange force of the magnetic layers, in the medium shown in Fig. 13;
Fig. 15 ïs a chart showing the temperature characteristic of the effective bias magnetic field of the magnetic layers in the medium shown in Fig. 13;

Figs. 16 and 17 are charts showing the temperature characteristic of the magnetic field required for orienting the magnetization of the adjusting layer in the perpendicular direction;
Fig. 18 is a chart showing the temperature 25 characteristic of the coercive force of the magnetic layers in the medium shown in Fig. 3;
Fig. 19 is a chart showing the temperature 1 characteristic of the coercive force and exchange force of the magnetic layers in the medium shown in Fig. 3;
Figs. 20 and 21 are schematic views showing states of magnetization in the recording process 5 utilizing the compensation temperature;
Figs. 22 and 23 are charts showing the temperature characteristic of the coercive force, when the magnetic layer has the compensation temperature between the room temperature and the Curie temperature;
Figs. 24, 25 and 26 are schematic cross-sectional views of embodiments of the magnetooptical recording medium of the present invention provided with a protective layer;
Figs. 27 and 28 are schematic cross-15 sectional views of embodiments of the magnetoopticalrecording medium of the present invention utilizing triple-layered magnetic film;
Figs. 29, 31 to 33 are charts showing states of magnetization in the recording process utilizing 20 the medium shown in Fig. 27;
Figs. 30 and 34 are charts showing the temperature characteristic of the coercive force of the magnetic layers in the medium shown in Fig. 27;
Figs. 35 and 36 are schematic cross-25 sectional views of embodiments of the magnetoopticalrecording medium of the present invention utilizing quadraple-layered magnetic film;

- 10 - 1~ 100 ~

1 Fig. 37 is a schematic view showing states of magnetization in the recording process utilizing the medium shown in Fig. 35;
Fig. 38 is a chart showing the temperature 5 characteristic of the coercive force of the magnetic layers in the medium shown in Fig. 35;
Figs. 39, 4l, 42 and 43 are charts showing states of magnetization in the recording and erasing processes utilizing the medium shown in Fig. 3;

Fig. 40 is a chart showing the relation between the erasing laser power and the residual signal after erasure in the medium shown in Fig. 3;
Fig. 44 is a chart showing the relation between the recording laser power and the reproduced 15 C/N ratio in the medium shown in Fig. 3; and Figs. 45 and 46 are schematic views showing variations of the magnetooptical recording apparatus utilizing the medium of the present invention.

Now the present invention will be clarified in detail by embodiments thereof shown in the attached drawings.
Fig. 3 is a schematic cross-sectional view 25 of an embodiment of the magnetooptical recording medium of the present invention. Said medium is composed of a translucent substrate l provided with 0 0; ~

1 guide grooves in advance (called "pregrooved"), and a a first magnetic layer 2 and a second magnetic layer 3 laminated thereon. The first magnetic layer 2 has a lower Curie point (TL) and a higher coercive force (HH), while the second magnetic layer 3 has a higher Curie point (TH) and a lower coercive force (HL).
The terms "higherN and "lower" are defined through relative comparison of the two magnetic layers, the comparison of the coercive force being at the room temperature. This comparison is more detailedly illustrated in Fig. 4. It is generally desirable that the first magnetic layer 2 has TL in a range of 70 - 180 ~C and HH in a range of 3 - l0 KOe, and the second magnetic layer 3 has TH in a range of 150 - 400 ~C and HL in a range of 0.5 - 2 KOe.
The thickness, coercive force, saturated magnetization and magnetic wall energy of said magnetic layers 2, 3 are so selected that the two states of the finally recorded bit can exist in stable manner. The magnetic layers 2, 3 may be exchange-coupled or magnetostatically coupled. However, an exchange-coupling is preferable in consideration of the magnitude of the effective bias magnetic field at the recording and the stability of the recorded binary bit.
In the magnetooptical recording medium of the present invention, the first magnetic layer ... "",,. , . , ,,, ., ,. ~ . ~ ",.. . . .. ...

- 12 - ~ 0 1 2 is principally related to the reproduction. The magnetooptical effect exhibited by said first magnetic layer 2 is principally utilized in the reproduction, while the second magnetic layer 3 plays an important role in the recording.
On the other hand, in the conventional exchange-coupled double-layered magnetic film mentioned above, the magnetic layer with a lower Curie point and a higher coercive force is principally related to the recording, and the magnetic layer with a higher Curie point and a lower coercive force is principally related to the reproduction. In such conventional exchange-coupled double-layered film, there stands a relationship:

HH > ~ w > H (1) 2Msh L

among the saturation magnetization Ms of the latter magnetic layer, film thickness h and magnetic wall energy aw between two layers.
On the other hand, in the exchange-coupled double-layered film of the recording medium of the present invention, there is required a relation:

H L 2MSh (2) among the saturation magnetization Ms of the second - 13 - 1 ~ 4 0 a-;~

1 magnetic layer, film thickness h and magnetic wall energy ~w between two layers.
This condition will be detailed later but in simpler terms is required for stabilizing the state of magnetization of the bit finally formed by recording, as shown by 4f in Fig. 5. Consequently the effective bias magnetic field, thickness, coercive force, satura-tion magnetization, magnetic wall energy etc. of the magnetic layers 2, 3 can be so determined as to satisfy the above-mentioned relation.
Each magnetic layer can be composed of a substance exhibiting a vertical magnetic anisotropy and a magnetooptical effect, preferably an amorphous magnetic alloy of a rare-earth element and a transi-tion metal element such as GdCo, GdFe, TbFe, DyFe,GdTbFe, TbDyFe, GdFeCo, TbFeCo, GdTbCo or GdTbFeCo.
In the following there will be given as explanation on a recording process utilizing the above-explained magnetooptical recording medium, while making reference to Fig. 5, showing the states of magnetization of the magnetic layers 2, 3 in the steps of the recording process, and Fig. 6 schemati-cally showing a recording apparatus. Prior to the recording, the stable directions of magnetization of the magnetic layers 2, 3 may be mutually same or opposite. Fig. 5 shows a case in which said stable directions of magnetization are mutually same.

l In Fig. 6, it is assumed that a part of a magnetooptical disk 9 of the above-explained structure has an initial magnetization as shown by 4a in Fig. 5.
The magnetooptical disk 9, being rotated by a spindle motor, passes the position of a magnetic field generating unit 8, generating a magnetic field of which intensity is selected at a suitable value between the coercive forces of the magnetic layers 2, 3 (magnetic field being directed upwards in the present embodiment), whereby, as shown by 4b in Fig. 5, the second magnetic layer 3 is uniformly magnetized while the first magnetic layer 2 retains the initial magnetization state.
The rotated magnetooptical disk 9, in passing the position of a record/reproducing head 5, is irradiated by a laser beam having two laser power levels according to the signal from a recording signal generator 6. The first laser power is enough for heating the disk to a temperature close to the Curie point of the first magnetic layer 2, while the second laser power is enough for heating the disk to a temperature close to the Curie point of the second magnetic layer 3. More specifically, referring to Fig. 4 showing the relation between the temperature and the coercive forces of the magnetic layers 2, 3, the first laser power can heat the disk close to TL
while the second laser power can heat it close to TH.

- 15 - 13~00.8 1 The first laser power heats the first magnetic layer 2 close to the Curie point thereof, but the second magnetic layer 3 has a coercive force capable of stably maintaining the bit at this temperature.
Thus, through a suitable selection of the recording bias magnetic field, a record bit shown in 4c can be obtained, as a first preliminary recording, from either state in 4b in Fig. 5.
The suitable selection of the bias magnetic field has the following meaning. In said first pre-liminary recording, such bias magnetic field is essentially unnecessary, since the first magnetic layer receives a force (exchange force) to arrange the magnetization in a direction stable to the direction of magnetization of the second magnetic layer 3, said directions being same in this case.
However said bias magnetic field is provided, in a second preliminary recording to be explained later, in a direction to assist the magnetic inversion of the second magnetic layer 3, namely in a direction prohibiting the first preliminary recording, and it is convenient to maintain said bias magnetic field in a same intensity and a same direction, both in the first and second preliminary recordings.

In consideration of the foregoing, the bias magnetic field is preferably selected at a minimum necessary intensity required for the second - 16 - 1 3 IQf) )8 1 preliminary recording to be explained in the following, and such selection corresponds to the suitable selec-tion mentioned above.
In the following there will be explained the second preliminary recording. This is achieved by heating the disk with the second laser power close to the Curie point of the second magnetic layer 3, whereby the direction of magnetization of the second magnetic layer 3 is inverted by the bias magnetic field selected as explained above, and the direction of magnetization of the first magnetic layer 2 is also arranged in a stable direction (same direction in the present case) with respect to the second magnetic layer 3. In this manner a bit as shown by 4d in Fig. 5 can be formed from either state shown in 4b.
Thus each area of the magnetooptical disk can have a preliminary record of the state 4c or 4d in Fig. 5, respectively by the first or second laser power corresponding to the input signal.
Then the magnetooptical disk 9 is further rotated and passes the position of the magnetic field generating unit 8, generating a magnetic field of which intensity is selected between the coercive forces of the magnetic layers 2, 3 as explained before, whereby the record bit 4c remains unchanged and assumes a final record state 4e, while the record - 17 - 13~

1 bit 4d assumes another final record state 4f as the result of magnetic inversion of the second magnetic layer 3.
In order that the record 4f can stably exist, there is required the aforementioned relationship (1):

~ w 2Msh L H

among the saturation magnetization Ms ~f the second magnetic layer 3, film thickness h and magnetic wall energy a w between the magnetic layers 2, 3. ~W/2Msh indicates the magnitude of the exchange force received by the second magnetic layer, or represents the magnitude of a magnetic field acting to rearrange the magnetiza-tion of the second magnetic layer 3 in a stable direction(same direction in the present case) with respect to the direction of magnetization of the first magnetic layer 2. Therefore, in order that the second magnetic layer 3 can retain its magnetization unchanged against said magnetic field, said layer should have a coercive force HL larger than the magnitude of said magnetic field tHL > aW/2Msh).
Stated differently, in order that the bit can stably exist, following relations are required among the coercive forces HH, HL of the first and second magnetic layers and the effective bias magnetic Heff~ HLeff of said layers:

,,, . ,~ , ., ., . 11 " ., ~, , . ~ , .

1 H Heff ~ L Leff ( ) HH - HHeff ~ HL + Leff These relations will be explained in more detail in relation to Figs. 7A to 7D. Fig. 7A is a chart showing the B-H loop, or the relation between the external magnetic field, in abscissa, applied to the first magnetic layer formed as a single layer, and the magnitude of the magnetization in said layer in ordinate. The chart indicates that, when the magnetic field is intensified in a direction of (+), the magnetization is aligned in a direction (+) or a direction (~) at an intensity HH, and, when the magnetic field is intensified in a direction (-), the magnetization is aligned in a direction (-) or (~) at an intensity -HH. Fig. 7B shows a similar B-H loop for the second magnetic layer formed as a single layer.

Fig. 7C shows a B-H loop of the first magnetic layer when the first and second magnetic layers are superposed with exchange-coupling and when said second magnetic layer is magnetized upward.
In contrast to the case of single layer shown in Fig. 7A, an effective bias magnetic field HHeff is applied, facilitating to align the magnetization of the first magnetic layer with that of the second 0 ;; ~

1 magnetic layer.
In order that the record bit 4f shown in Fig.
5 can stably exist, the state of a point A, where the direction of magnetization of the first magnetic layer is opposite to that of the second magnetic layer under a zero external field should be stable and should not transform to the state of a point B
wherein said directions of magnetization of the magnetic layers are mutually same. For this reason there is required a condition HH ~ HH ff > ~
Fig. 7D shows a similar B-H loop of the second magnetic layer when the first and second magnetic layers are superposed with exchange-coupling and when said first magnetic layer is magnetized up-wards, In contrast to the case of single layershown in Fig. 7B, an effective bias magnetic field HLeff is applied, facilitating to align the magnetiza-tion of the second magnetic layer with that of the first magnetic layer. In order that the record bit 4f in Fig. 5 can stably exist, the state of a point A, where the direction of magnetization of the first magnetic field is opposite to that of the second magnetic field under a zero external field should be stable and should not transform to the state of a point B wherein said directions are mutually same.
For this reason there is required a condition 1 3 ~-~ 11 0 ~ ~

1 L Leff Either in the first or second magnetic layer, an inversion of magnetization from a stable state to an unstable state requires a magnetic field equal to the coercive force of the magnetic layer plus the exchange force, since such inversion has to be made against the exchange force acting on said layer.
On the other hand, an inversion from an unstable state to a stable state requires a magnetic field equal to the coercive force of the magnetic layer minus the exchange force, since the exchange force facilitates the inversion in this case.
Therefore, in order that the magnetization of the first magnetic layer is not inverted in the lS magnetic field generating unit 8 and the magnetiza-tion of the second magnetic layer is aligned to the direction of the magnetic field of said unit in any combination of the magnetized states, the external field B should be adjusted to an internal level if there stands a relation:
HH - HHeff ~ HL + HLeff This is because the magnetic field required for inverting the magnetization of the second magnetic layer is larger than HL + HLeff when the first and second magnetic layers are in a stable state, and because the magnetic field not inducing the inversion of magnetization of the first magnetic layer should - 21 - 1 310~

1 be smaller than HH ~ HHeff when the first and second magnetic layers are in a stable state.
An overwriting operation is therefore rendered possible, since the record bits 4e, 4f do not rely on the state prior to recording but only on the level of laser power at the recording. The record bits 4e, 4f can be reproduced by irradiation with a reproducing laser beam and processing with a signal reproducing unit 7.
Fig. 8 is a schematic view showing a more detailed embodiment of the recording apparatus shown in Fig. 6. In Fig. 8 there are shown a magnetooptical disk or recording medium 11 of a structure as shown in Fig. 3; a spindle motor 12 for rotating said disk 11;
a clamper 13 for fixing the disk 11 on the rotating shaft of the motor 12; and an optical head 14 for projecting a light beam 15 onto the disk 11. Said optical head 14 is provided with a laser light source 16 composed for example of a semiconductor laser; a collimating lens 17; a beam splitter 18; an objective lens 19; a sensor lens 20; an analyzer 26 and a photo-detector 21, and is radially moved by an unrepresented mechanism. Also the objective lens 19 moves in the axial direction and a direction perpendicular thereto to achieve so-called auto tracking (AT) and auto focusing (AF), according to control signals detected by the photodetector in an already known manner.

- 22 - 13~

1 The laser light source 16 is driven by a laser driver circuit 22 and emits a light beam 15 modulated in intensity between two non-zero values, according to the recording information entered from an input terminal 23, as will be explained later.
In a position opposed to the optical head ~4 across the disk ll, there is provided first magnetic field generating means 24 to apply a bias magnetic field of a predetermined direction to an area of the 10 disk 11 irradiated by the light beam 15. Also at a position distance by 180~ in the rotating direction of the disk 11, there is provided second magnetic field generating means 2S for applying a bias magnetic field of a direction opposite to said predetermined direction. Said first and second magnetic field generating means may be composed of electromagnets, but the use of permanent magnets is preferable for simplifying the apparatus and reducing the cost thereof, since the direction of magnetic fields need not be switched in the present invention.
In the following explained is the process of information recording with the apparatus shown in Fig. 8. The light beam lS from the laser unit 16 is modulated, as shown in Fig. 9, between two non-zero levels P1 and P2, corresponding to binary recording signals "0" and "1". The light beam of the level Pl has an energy for heating the magnetic layers 2, 3 - 23 -~ 13 iO 0 ;~

of the medium shown in Fig. 3 to the Curie temperature TL of the first magnetic layer, while that of the level P2 has an energy for heating said magnetic layers to the Curie temperature TH of the second magnetic layer. By the irradiation with such modulated beam, the magnetic layers of the disk ll undergo changes in magnetization as shown in 10a - 10h in Fig. lOA, thereby recording information.
In Fig. 10A, arrows in the first and second magnetic layers 2, 3 indicate the directions of magne-tization, and the length of arrow indicates the magnitude of coercive force. An upward magnetized area, shown in 10a, is subjected to an upward bias magnetic field of -B2 by the second magnetic field generating means 25 as shown in 10b, and is then moved to the position of the optical head 14 by the rotation of the disk for irradiation with the modulated light beam 15. In response to the light beam of the power P1, the first magnetic layer 2 alone reduces its coercive force as shown in 10c but retains the upward magnetization as shown in 10d due to the magnetic interaction of the second magnetic layer 3, despite of the application of a bias magnetic field B1 by the first magnetic field generating means 24. On the other hand, in response to the light beam of the power P2, the magnetization is inverted downwards as shown in 10h by the application - 24 - 1 ~ ~ 0 0~5 1 of the bias field B1.
On the other hand, in a downward magnetized area as shown in 10e, the second magnetic layer 3 alone inverts the magnetization by the application of a bias magnetic field -B2. Then, in response to the light beam of the level Pl, the first magnetic layer 2 reduces the coercive force of the first magnetic layer 2 as shown in 10c, and inverts the magnetiza-tion upwards as shown in 10d by the magnetic inter-action of the second magnetic layer 3. Also inresponse to the light beam of the power P2, both magnetic layer reduce the coercive force as shown in 10g, and retain the downward magnetization as shown in 10h by the application of the bias magnetic field B1, As explained in the foregoing, the apparatus shown in Fig. 8 can determine the direction of magne-tization solely according to the change in the power of the light beam, regardless of the initial magneti-zation of the magnetic film. Consequently the alreadyrecorded information need not be erased but can be rewritten by direct overwriting. The recorded information can be reproduced in conventional manner, by detecting the direction of magnetization of the first magnetic layer utilizing the magnetooptical effect. As an example, in the apparatus shown in Fig. 8, the recorded information can be reproduced - 25 ~ tl~ ?8 1 as electrical signals, by causing the laser unit 16 to continuously emit a light beam of a power insuffi-cient for heating the disk to the Curie point of the second magnetic layer and receiving the reflected light from the disk 11 by the photodetector 21 through the analyzer 26.
In the apparatus shown in Fig. 8, the magnetic fields B1, B2 applied to the magnetic layers respectively by the first and second magnetic field generating means are so selected as to satisfy the following relations:
H ~ B2 < H (5) L ' o (6) H ' < H = Bl (7) H ' L (8) wherein HH' and HL' are coercive forces of the first and second magnetic layers at a temperature TL, HH"
and HL" are coercive forces at a temperature TH, and Ho is the magnitude of the magnetic interaction between two layers.
In the foregoing description it is assumed the first and second magnetic layers 2, 3 are stable when the magnetizations thereof are in a same direc-tion, but a similar process is applicable also when said magnetizations are stable when they are mutually oppositely directed. Fig. lOB illustrates states of magnetization in the recording process of such case, . ,, . , ... . . . ., , ~" ,. . . . . .

- 26 - ~3'~ 05 l wherein 40a to 40f respectively correspond to 4a to 4f in Fig. 5.

Example 1 A polycarbonate substrate with pregrooves and preformat signals was set in a sputtering apparatus with three targets, and was rotated at a distance of 10 cm from the target.
A ZnS protective layer of 1000 A in thickness was obtained by sputtering from a first target in argon gas, with a sputtering speed of 100 A/min., and a sputtering pressure of 5 x 10 3 Torr. Then a TbFe alloy was sputtered from a second target, in argon gas, with a sputtering speed of 100 A/min. and a sputtering pressure of 5 x 10 Torr to obtain a first magnetic layer of Tb18Fe82 with a thickness of 500 A, TL of about 140 ~C and HH of about 10 KOe.
Then a TbFeCo alloy was sputtered in argon gas at a sputtering pressure of 5 x 10 Torr to obtain a second magnetic layer of Tb23Fe60Co17 with a thick-ness of 500 A, TH of ca. 250 ~C and HL of ca. 1 KOe.
Subsequently a ZnS protective layer of 3000 A in thickness was obtained by sputtering from the first target in argon gas, with a sputtering speed of 100 A/min. and a sputtering pressure of 5 x 10 3 Torr.
After these layer formations, the l~Od5~

above-mentioned substrate was adhered to a polycarbonate substrate with hot-melt adhesive material to obtain a magnetooptical disk. Said disk was mounted on a record/
reproducing apparatus and was made to pass through, with a linear speed of 8 m/sec., the second magnetic field generating means for applying a field of 1500 Oe to the disk. The recording was conducted with a laser beam of a wavelength of 830 nm, concentrated to ca. 1 ~m and modulated in two levels of 4 and 8 mW, with a duty ratio of 50 ~ and a frequency of 2MHz.
The bias magnetic field was 100 Oe. Binary signals could be reproduced by irradiation with a laser beam of 1.5 mW.
The above-explained experiment was conducted on a magnetooptical disk already recorded over the entire surface. The previously recorded signal components were not detected in the reproduction, and the possibility of overwriting was thus confirmed.
Example 2 A magnetooptical disk was prepared in the same manner as in the Example 1, except that the second magnetic layer was composed of Tb23Fe70Co7 with TH = 200 ~C, HL = ca. 1 KOe and HL ff = ca.
300 Oe, and was subjected to recording and reproduc-tion in the same manner as in the Example 1 except the use of a magnetic field generating field of ca.
2.5 KOe to obtain similar results as those in the - 28 - ~ 0 o.i 1 Example 1.
The magnitude of HHeff in this case will be explained in the following in relation to Figs. 7C and 7D.
Since HH - HH ff was larger than HL ~ HL ff' when the first and second magnetic layers were magne-tized in a same direction and subjected then to an inverting magnetic field, the magnetization of the second magnetic layer was inverted at HL ~ HL ff =
0.7 KOe so that HLeff could not be measured.
However, based on the condition HH ~ HHeff ~ HL ~ HL ff~ a conclusion 4.3 KOe > HH ff from the conditions HH = 5 KOe, HL = 1 KOe and HL ff = 0 3 KOe.
Also the HHeff was measured as ca. 1 KOe in an experiment in which a first magnetic TbFe layer of same composition and thickness was superposed with a second magnetic TbFeCo layer with a modified composition to increase HL HLeff It was confirmed, for the above-mentioned coercive forces and exchange force, that the first and second magnetic layers satisfied the condition:
HH ~ HHeff ~ HL + HLeff by which the second magnetic layer alone is magnetized in the direction of the magnetic field from the magnetic field generating unit 24.
Example 3 A polycarbonate disk-shaped substrate with 1 pregrooves and preformat signals was set in a sputter-ing apparatus with three targets, and was rotated at a distance of 10 cm from the target.
A protective SiC layer of 700 A in thick-ness was obtained by sputtering from a first targetin argon gas, with a sputtering speed of 100 A/min.
and a sputtering pressure of 3 x 10 Torr. Then a GdTbFe alloy was sputtered from a second target, in argon gas, with a sputtering speed of 50 A/min. and a sputtering pressure of 3 x 10 3 Torr to obtain a first magnetic layer of Tb8Gdl2Fe80 with a thickness of 200 A, TL = ca. 160 ~C and HH = ca. 8 KOe.
Then a TbFeCo Cu alloy was sputtered in argon gas at a sputtering pressure of 3 x 10 3 Torr to obtain a second magnetic layer of Tb23Fe50Co15Cul2 with a thickness of 400 A, TH = ca. 180 ~C and HL =
ca. 1 KOe.
Subsequently a Si3N4 protective layer of 1200 A in thickness was obtained by sputtering from the first target in argon gas, with a sputtering speed of 70 A/min. and a sputtering pressure of 3 x 10 Torr.
After these layer formations, the above-mentioned substrate was adhered to a polycarbonate plate with hot-melt adhesive material to obtain a magnetooptical disk. Said disk was mounted on a record/reproducing apparatus and was made to pass 1~oo~)~

1 through, a with a linear speed of 8 m/sec., the second magnetic field generating means for applying a field of 1500 Oe to the disk. The recording was then con-ducted with a laser beam of a wavelength of 830 nm, concentrated to ca. 1~ m and modulated in two levels of 4 and 8 mW, with a duty ratio of 50 ~ and a frequency of 2 MHz. The bias field at the irradiated area was 100 Oe. Binary signals could be reproduced by irradiation with a laser beam of 1.5 mW.
The above-explained experiment was conducted on a magnetooptical disk already recorded over the entire surface. The previously recorded signal com-ponents were not detected in the reproduction, and the possibility of overwriting was thus confirmed.
In case the apparatus shown in Fig. 6 employs a permanent magnet as the magnetic field generating unit 8, the magnetooptical recording medium of the present invention is constantly exposed, in the recording and reproducing operations, to a magnetic field generated by said magnetic field generating unit. Even in the reproduction, the medium is subjected to a laser beam irradiation of an energy of ca. 1/3 to 1/10 of the energy at the recording, and may therefore reach a temperature of about 70 ~C at maximum in passing the magnetic field generating unit. Thus the magnetization of the first magnetic layer 2 may become inverted in repeated - 31 - 13 10 0~

l reproductions, for example of 10 times. In order to avoid such inversion of magnetization, it is pre-ferable, as shown in the following examples, to select the coercive force HH of the first magnetic layer and the magnetic field B applied by the magnetic field generating unit in such a manner as to satisfy a relation 0.2 x HH ~ 0-3 > B (KOe~, and, more pre-fearbly a condition HH ~ 1.5 KOe. A further preferred condition is HH ~ 5 KOe as will be explained in the following.
Example 4 Samples of magnetooptical disk 4-1 to 4-12 were prepared in the same process as in the Example 1, with the same film thickness and the same structure except that the composition, coercive force HH and Curie point TL of the first magnetic layer 2 were changed.
Each of the disks of the Examples 1 and 4-1 to 4-12 was mounted on a record/reproducing apparatus and was made to pass through, with a linear speed of 8 m/sec., the second magnetic field generating means for applying a field of 2 KOe.
The recording was conducted with a laser beam of a wavelength of 830 nm, concentrated to ca. 1 ~m and modulated in two levels of 4 and 8 mW, with a duty ratio of 50 ~ and a frequency of 2 MHz. The bias magnetic field at the recording head was 100 Oe.

- 32 - 13 1~0 ;~

1 Then investigated was the change of the noise component in the reproduced signal, after 10 reproductions from a same track by irradiation with a laser beam of 1.0 mW and with varied magnetic field B generated by the magnetic field generating unit 8.
Subse~uently the temperature inside the apparatus was set at 30~, 45~ and 60~C and the inten-sity of magnetic field causing the increase of noise component in the reproduced signal was determined for each temperature, as summarized in Tab. 1.

1 c ~ ~ 1) f~

a ~ U~
o o a) :n u~
~ ~) t~ 1 0 0 0 ~ ~-1 0 ~ I O O

~ 0 m ~

o ~ o _ ~
a~
a) ~ u~
~ ~
m ~

U~
~rt ~: O
P
E~ ~ ~ ,a . . . . . . . . .
O
~ ~ U
15 ~ ~: o m ~, r1 ~) ~ O O O O O O O O O U~ O O O
--o ~t~
U~ U U O
~ ~ ~r U o o d' U
~ 0 CD 0~ ~0 00 0 0 a~ a) ~ v a~
4 4 ~ ~4 L ~1 a~
L ~ 4 V V V V V t4 0 ~ ~ ~ ~r ~r o o o o o u~
u) ~d Q P S~ ,D,D ~ D ~ D Q 5~ P D

~ o I
X

- 34 ~ l~ d;~

1 Tab. 1 indicates that the magnetic field B
inducing the increase of noise component in the repro-duced signal becomes smaller as the coercive force HH
of the first magnetic layer becomes smaller, or as the temperature inside the apparatus becomes higher, and that this relationship is not affected by the composi-tion of the first magnetic layer.
Fig. 11 illustrates the relation of HH value and the B value inducing the noise increase for these samples, based on the above-explained results.
From these relationships it is apparent that, in order to prevent the noise increase up to 60 ~C, HH and B should at least satisfy a relation:
0.2 x HH ~ 0-3 ~ B.

It is also apparent that the increase in noise takes placed even at a small value of B unless HH is at least equal to 1.5 KOe.
Since the temperature inside the apparatus does not exceed 60 ~C in practice, the increase in noise can be prevented even after prolonged repro-duction if the above-mentioned conditions are satisfied.
However, in addition to the foregoing conditions, it is preferable to maintain HH at least equal to 5 KOe for the following two reasons:
(i) The minimum necessary value of B is equal to the sum of the coercive force HL of the _ 35 ~ 0 0 )~

1 second magnetic layer and the magnetic field acting on the second magnetic layer as the result of the exchange force, and is considered in the order of 1 to 2 KOe in practice. Thus, it will be apparent from Fig. 11 that the value of HH should be equal to or higher than 5 KOe in order to maintain the noises at a low level in continuous reproduction at a temperature up to 60 ~C and with the magnetic field B at 1 - 2 KOe.
(ii) Fig. 11 indicates the magnitude of the magnetic field B and the value of HH required for suppressing the noise at the selected value of B at each temperature inside the apparatus. For example, at a temperature 60 ~C and in a range HH ~ 5 KOe, there is obtained a relation HH = 5(B + 0.3). Thus, for an increase ~ B of the magnetic field B generated by the magnetic field generating unit, there is required a corresponding increase 5 x A B in the value of HH.
However, in a range of HH exceeding 5 KOe, the increase required for HH corresponding to a certain increase in the magnetic field B at a temperature of 30~ - 60 ~C is smaller than that requlred in a range of HH below 5 KOe.
As explained in the foregoing, the magneto-optical recording medium of the present invention is required to satisfy the aforementioned relation (1):

- 36 - 13~n~3;-~

1 HH ~ HL ~ 2M h For this purpose it is also very effective to adjust the magnetic wall energy between the magnetic layers.
The adjustment of the magnetic wall energy can be achieved by following methods:
( I ) adjustments of the composition of the magnetic layers;
( II ) addition of a predetermined step in the preparation of the medium; and (III) formation of an adjusting layer for the magnetic wall energy, between the magnetic layers;
which will be explained further in the following.

[I. Adjustment of the composition of the magnetic layers]
In a magnetooptical recording medium of the structure as shown in Fig. 3, the magnetic wall energy can be reproducibly reduced by forming one of the first and second magnetic layer 2, 3 with a composi-tion richer in transition metals compared with the compensation composition, and forming the other layer with a composition richer in rare-earth elements.
Such example is shown in the following.

Example 5 A disk-shaped polycarbonate substrate with pregrooves and preformat signals was set in a sputtering _ 37 ~ 0~~i~

1 apparatus with three targets, and was rotated at a distance of 10 cm from the target.
A protective Si3N4 layer of 600 A in thick-ness was obtained by sputtering from a first target in argon gas, with a sputtering speed of 40 A/min.
and a sputtering pressure of 5 x 10 Torr. Then a TbFe alloy was sputtered from a second target, in argon gas, with a sputtering speed of 100 A/min.
and a sputtering pressure of 5 x 10 Torr to obtain a first magnetic layer of Tb3Gd16Fe81, richer in Fe with respect to the compensation composition, with a thickness of 400 A, TL = ca. 155 ~C and HH = ca. 8 KOe.
Then a TbFeCo alloy was sputtered in argon at a sputtering pressure of 5 x 10 3 Torr to obtain a second magnetic layer of Tb10Dy13Fe60Co17~ richer in Tb and Dy with respect to the compensation com-position, with a thickness of 300 A, TH = ca. 200 ~C
and HL = ca. 1 KOe.

Subsequently a Si3N4 protective layer of 1500 A in thickness was obtained by sputtering from the first target in argon gas, with a sputtering speed of 40 A/min. and a sputtering pressure of 5 x 10 3 Torr.

After these layer formations, the above-mentioned substrate was adhered to a polycarbonate plate with hot-melt adhesive material to obtain a - 38 ~ 0'i 1 magnetooptical disk. Said disk was mounted on a record/reproducing apparatus and was made to pass through, with a linear speed of 8 m/sec., a magnetic field generating unit for applying a field of 2.5 KOe.
The recording was conducted with a laser beam of a wavelength of 830 nm, concentrated to ca. 1 ~m and modulated in two levels of 4 and 8 mW, with a duty ratio of 50 % and a frequency of 2 MHz. The bias magnetic field at the irradiated area was 100 Oe, in a direction to invert the magnetization of the second magnetic layer. Binary signals could be reproduced by irradiation with a laser beam of 1.5 mW.
The above-explained experiment was conducted also on a magnetooptical disk already recorded over the entire surface. The previously recorded signal components were not detected in the reproduction, and the possibility of overwriting was thus confirmed.
Example 6 In the Example 5, the first magnetic layer was richer in Fe while the second magnetic layer was richer in Tb and Dy in comparison with the compensa-tion composition.
In the present example, there were prepared and evaluated samples of the magnetooptical disk, in which the first and second magnetic layers had same coercive forces as explained above, are composed of a combination of a composition richer in transition - 39 ~

1 metals such as Fe and a composition richer in rare earth elements such as Tb or Dy.
A coercive force of 8 KOe for the first magnetic layer is achieved, in a transition metal-rich composition, by Tb3Gd16Fe81, or, in a rare earth-rich composition, by Tb3 3Gd17 7Fe77. Also a coercive force of 1 KOe for the second magnetic layer is achieved, in a transition metal-rich composition Y 7 4DY9.6Fe64~7col8 3~ or, in a rare earth_rich cOmposition~ by TbloDyl3Fe6ocol7 Samples listed in Tab. 2 were prepared by selecting a transition metal-rich composition or a rare earth-rich composition mentioned above for the first and second magnetic layer, and selecting other materials and layer thicknesses same as those in the Example 5. These samples were then subjected to the record/reproducing experiment as in the Example 5.
The results are shown in Tab. 2.
"TM" and "RE" respectively show a composi-tion richer in the transition metals and a compositionricher in the rare earth elements, compared to the compensation composition.

, ~ ~. ,, , .. , ,., , . , " . .. .. .

a~
U~
JJ
P ~,~
U~
o . + + +l +l S~ P ~
o O ~:
.,~ .
~a G) u~
~ + +

~ P~

O O

h ~ O ~ ~r O
~ ~1 0a) a~ o P n~

a ~ o a ~ d' a O ~ ~ O, c: ~ P P ~ p ,~
CO ~ o~

~ o ~ ~ ~ ~
P P ~ P

X

1 "+" indicates that the record bits are stable in the absence of external magnetic field and provide satis-factory reproduction signals; "+" indicates that the record bits are partially inverted or the reproduction signals are of insufficient quality; and "-" indicates that the record bits are unstable.
These results indicate that stable record bits are obtained only when one of the first and second magnetic layers is composed of the transition metal-rich composition and the other is composed of the rareearth-rich composition.
[II. Addition of a predetermined step in the preparation of the medium]
In the preparation of the magnetooptical recording medium as shown in Fig. 3, a medium satisfying the afore-mentioned relation (1) can be easily obtained by adding one of following steps after the formation of the first magnetic layer and before the formation of the second magnetic layer:

(A) a step of standing in an atmosphere of remaining gas or inert gas at 7 x 10 7 Torr for 5 minutes of longer;
(B) a step of standing in an atmosphere with a partial pressure, at least equal to 2 x 10 Torr, of a substance capable of reacting with a constituent element of the first or second magnetic layer - 42 - 13'~

1 or being chemically absorbed by said element; or (C) a step of exposure to a plasma atmosphere of inert gas or a substance capable of reacting with a constituent element of the first or second magnetic layer or being chemically absorbed by said element.
The first and second magnetic layers can be formed by sputtering, or evaporation for example with electron beam heating.
Examples of the above-mentioned remaining gas are H2O, ~2~ H2, N2 and low-molecular compounds consisting of C, H, N and O, and examples of the inert gas are Ar, He and Ne.

Examples of the gas capable of reacting with the constituent element of the first or second magnetic layer or being chemically absorbed by said element are H2O~ ~2~ H2' N2' 2 ~ 2 4 In the usual manufacturing process for the medium, after the formation of the first magnetic layer, the formation of the second magnetic layer is conducted immediately (for example within 1 minute) in a clean high-vacuum atmosphere. However the addition of one of the steps (A) - (C) modifies the exchange force, coercive force or stability of the magnetic layers, and a reproducible recording characteristic can be obtained through precise - 43 - 13~00.~

1 control of the process conditions, process time etc.
The effects of the steps (A) - (C) will be verified in the following examples.
Example 7 A pregrooved and proformatted polycarbonate disk substrate was set in a sputtering apparatus with three targets, and was rotated at a distance of 10 cm from the target.
The apparatus was evacuated to 1 x 10 Torr, and a protective SiO layer of 1000 A in thick-ness was obtained by sputtering from a first target in argon gas, with a sputtering speed of 100 A/min.
and a sputtering pressure of 5 x 10 Torr. Then a TbFe alloy was sputtered from a second target, in argon gas, with a sputtering speed of 100 A/min. and a sputtering pressure of 5 x 10 Torr to obtain a first magnetic layer composed of Tbl8Fe82 with a thickness of 300 A, TL = ca. 140 ~C and HH = ca. 10 KOe. After the completion of sputtering, argon gas supply was continued for 30 minutes, with a pressure of 5 x 10 Torr in the sputtering chamber.
Then a TbFeCo alloy was sputtered in argon gas with a sputtering pressure of 5 x 10 3 Torr to obtain a second magnetic layer composed of Tb23Fe70Co7 with a thickness of 400 A, TH = ca. 200 ~C and HL =
ca. 1 KOe.
Subsequently a SiO protective layer of l~ ~lQ05~

1 2000 A in thickness was formed by sputtering from the first target in argon gas, with a sputtering speed of 100 A/min. and a sputtering pressure of 5 x 10 Torr.
After these layer formations, the above-mentioned substrate was adhered to a polycarbonateplate with hot-melt adhesive material to obtain a magnetooptical disk. Said disk was mounted on a record/reproducing apparatus and was made to pass through, with a linear speed of 8 m/sec., a unit for generating a magnetic field of 2.5 KOe. The recording was conducted with a laser beam of a wavelength of 830 nm, concentrated to ca. 1 ~m and modulated in two levels of 4 and 8 mW, with a duty ratio of 50 % and a frequency of 2 MHz. The bias magnetic field was 100 Oe. Binary signals could be reproduced by irradiation of a laser beam of 1.5 mW.
The above-explained experiment was repeated also on a magnetooptical disk already recorded over the entire surface. The previously recorded signal components were not detected in the reproduction, so that the possibility of overwriting was thus comfirmed.
Example 8 and Reference Example Samples of the magnetooptical disk were prepared in a process similar to that of the Example 7 but with varied conditions (atmosphere and pressure) in the step between the formations of the first and - 45 - 13'1~

1 second magnetic layers, as listed in Tab. 3. Those marked with * are examples of the present invention, and others are reference examples.
In examples 8-30 to 8-33, a disk electrode of 20 cm in diameter was placed at 5 cm from the poly-carbonate substrate, and a plasma treatment was conducted with a discharge power of 50 W, in the presence of various gasses listed in Tab. 3 at a pressure of 5 x 10 3 Torr in the sputtering chamber.

In examples 8-5 to 8-14, the main valve of the vacuum pump was suitably closed to vary the remaining gas atmosphere.
Each sample was evaluated for the stability of the record bits 4f, shown in Fig. 5, in the absence of external magnetic field, through the measurement of an external magnetic field inducing the inversion of magnetization in the magnetic layers. "+" and "-" respectively indicate that the record bits are stable or unstable.
Also each samples was tested for recording and reproduction in the same manner as in the Example 7. +" and "-" respectively indicate that the record-ing was satisfactorily or unsatisfactorily made.

- 46 - I3~

Tab. 3 Example Atmosphere Vacuum Time Sample evaluation (Torr) (min) Stability Record of bits 4f state * Ex. 7 argon gas 5 x 10 30 + +
8- l ' 1/4 - -* 8- 2 ~ 2 * 8- 3 " " 5 + +
8- 4 " " 15 + +
8- 5 remaining gas 1 x lO 6 1/4 lO 8- 6 " " 2 * 8- 7 ~ " 5 + +
* 8- 8 " " l5 + +
* 8- 9 " " 30 + +
8-lO remaining gas 3 x 10 l/4 l;~ 8-11 ' ' 2 * 8-12 " ~ 5 + +
* 8-13 " " 15 + +
* 8-14 " ~ 30 + +
* 8-15 oxygen gas 3 x 10 l/2 + +
* 8-16 " " 2 + +
~0 * 8-17 ~ " 5 * 8-18 " " 15 + +
* 8-19 " ~ 30 + +
* 8-20 nitrDgen gas 3 x 10 1/2 + +
25 * 8-21 " ~ 2 + +
* 8-22 " ~ 5 + +
* 8-~3 " " 15 + +

~, , . , , ,. , ,,,, ,, ~ , " ,, , , , . ,. ,, ~,, ,, " ,, , , , ", , .

- 4 7 - 1 3 1 ~ O ~j ~

Example Atmosphe re Vacuum Time Sample evaluation (Torr) (min) Stability Record of bits 4f state * 8-24 nit rogen gas 3 x 10 30 + +
* 8-25 hydrogen gas 3 x 10 1~2 - -5 * 8-26 " " 2 + +
* 8-27 ~ .. 5 +
* 8-28 " " 15 + +
* 8-29 " " 30 + +
* 8-30 argon plasma 3 x 10 1/12 + +

10 * 8-31 oxygen plasma " 1/12 + +
* 8-32 nitrogen plasma " 1/12 + +
* 8-33 hydrogen plasma " 1/12 + +

.. . . ~, . . .. .....

- 48 - 134~

l Results of the Example 7, Examples 8 and Reference Examples indicate that the recording by overwriting can be satisfactorily achieved by the magnetooptical recording medium prepared employing either one of the steps (A) to (C).
[ III. Formation of a magnetic wall energy adjusting layer between the magnetic layers]
Formation of an adjusting layer 41, as shown in Fig. 12, between the first and second magnetic layers 2, 3 allows to arbitrarily regulate the magnetic wall energy therebetween, thereby obtaining a medium satisfying the aforementioned condition (1).
More practically there may be employed a structure lS shown in Fig. 13, using a pregrooved substrate 1 and provided with protective layers 42, 43. In Figs. 12 and 13, same components as those in Fig. 3 are repre-sented by same numbers and will not be explained further.

Said adjusting layer 41 may be composed of a material not deteriorating the magnetic layers, for example Ti, Cr, Al, Ni, Fe, Co, rare earth element or a fluoride thereof.
The thickness of said adjusting layer 41 is suitably selected in consideration of the materials and thicknesses of the first and second magnetic layers, but is generally selected within a range of - 49 - ~ 3 4 ~, ~ ) 8 1 5 to 50 A.
Example 9 A pregrooved and preformatted polycarbonate disk substrate was set in a sputtering apparatus with ternary targets, and was rotated at a distance of 10 cm from the target.
A ZnS protective layer of 1000 A in thickness was formed by sputtering from a first target in argon gas, with a sputtering speed of 100 A/min. and a sputtering pressure of 5 x 10 Torr. Then a TbFe alloy was sputtered from a second target, in argon gas, with a sputtering speed of 100 A/min. and a sputtering pressure of 5 x 10 Torr to obtain a first magnetic layer composed of Tb18Fe82 with a thickness of 500 A, TL = ca. 140 ~C and HH = ca.
5 KOe.
Then a Co adjusting layer was formed with a thickness of 10 A, by sputtering in argon gas with a sputtering pressure of 5 x 10 3 Torr. Then TbFe and Co were simultaneously sputtered from second and third targets in argon gas with a sputtering pressure of 5 x 10 3 Torr to obtain a second magnetic layer 15Fe68Col7 with a thickness of ca 200 A, TH = ca. 250 ~C and HL = ca. 2 KOe.

Subsequently a ZnS protective layer of 3000 A in thickness was formed by sputtering from the first target in argon gas, with a sputtering - 50 ~ r~

1 speed of 100 A/min. and a sputtering pressure of 5 x Torr.
After these layer formations, the above-mentioned substrate was adhered to a polycarbonate plate with hot-melt adhesive material to obtain a magnetooptical disk. Said disk was mounted on a record/reproducing apparatus and was made to pass through, with a linear speed of 8 m/sec., a unit for generating a magnetic field of 2.5 KOe. The record-ing was then conducted with a laser beam of a wave-length of 830 nm, concentrated to ca. 1~ m and modulated in two levels of 4 and 8 mW, with a duty ratio of 50 % and a frequency of 2 MHz. The bias magnetic field was 100 Oe. Binary signals could be reproduced by irradiation with a laser beam of 1.5 mW.

0 .~ 8 1 The above-explained experiment was repeated with a magnetooptical disk already recorded over the entire surface. The previously recorded signal components were not detected in the reproduction, so that the possibility of overwriting was thus confirmed.
The adjusting layer was composed of Co in the foregoing embodiment, but it may also be composed of a magnetic material of which easy direction of magnetization is positioned longitudinally along the disk surface at room temperature but vertically to the disk surface at the recording temperature. The use of such material reduces the magnetic wall energy between the magnetic layers at room temperature and provides a larger exchange force to the magnetic layers at recording, thereby 1~ providing a magnetooptical recording medium enabling overwrite recording with a smaller bias magnetic field and superior in the stability of the record bits. Such structure will be shown in the following examples.
A sample for exchange force measurement was prepared by sputtering, on a slide glass, a first magnetic layer of Tbl8Fe82 of 500 A in thickness, then an adjusting layer of Fe or Tb25Fe70Co5 in various thicknesses, and a second magnetic layer of Tb22Fe70Co8 of 500 A in thick-ness. The first magnetic layer showed a coercive force of 12 KOe, with prevailing Fe sub-lattice magnetization, while the second magnetic layer showed a coercive force of 6 KOe, with prevailing Tb sub-lattice magnetization.

5' 13~0.ii3 1 Each sample was subjected to the measurement of the external magnetic field inducing the inversion of magnetization of the first and second magnetic layers, in a VSM and in the presence of an external magnetic S field. In a decreasing magnetic field, the samples showed an inversion of magnetization of the second magnetic layer into a stable (opposite) direction with respect to that of the first magnetic layer. The exchange force applied to the second magnetic layer was determined from such inversion-inducing magnetic field, as shown in Fig. 14, which indicates the exchange force on the second magnetic layer in ordinate and the thickness of the adjusting layer (Fe or TbFeCo) in abscissa.
As will be apparent from this chart, the exchange force is annulated even at a layer thickness of 70 A
with a Fe layer having a longitudinal easy direction of magnetization. On the other hand, in case of the Tb25Fe70Co5 with a coercive force of ca. 300 Oe and with a vertical easy direction of magnetization, parallel to those of the first and second magnetic layer, the exchange force is still effective even at a layer thickness of 500 A.
Therefore, the stability of record bits and the stable recording characteristic can be both obtained by forming, between the first and second magnetic layers, an adjusting layer of a material of which easy direction - 5.~ ~ 3 ~
..

l of magnetization is longitudinal at room temperature but is vertical at the recording temperature.
Temperature-dependent change of easy direction of magnetization is already known in substances showing S spin rearrangement. For example, DyCo5, reported by M.
Ohkoshi and H. Kobayashi in Physica, 86 - 88B (1977), p.195 - 196, exhibits a change of the easy direction of magnetization from longitudinal to vertical in a temperature range of 50~ - 100~C. Similar results are known in compounds in which Dy is replaced by another rare earth element such as Nd, Pr or Tb, or in which Co is replaced by another transition metal such as Fe or Ni. Also Tsushima reported, in Oyo Buturi, 45, 10 (1976), p.962 - 967, the spin rearrangement in rare earth la orthoferrites and rare earth orthochromites. A suitable modification of the composition of these substances allows to achieve a change of the easy direction of magnetization from longitudinal to vertical state in the recording temperature range.
Also it is already known that a thin magnetic film has to satisfy a condition:

H ' 4~M
k s in order to have a magnetization vertical to the film surface, wherein Ms is the saturated magnetization and Hk is the uniaxial anisotropic magnetic field in said vertical direction.
Therefore, in order that the adjusting layer 5 4 1 ~ .; 8 1 has the easy magnetization axis in the longitudinal direction at room temperature and in the vertical direction at the recording temperature range, it is desirable to select the Curie point of said adjusting layer in the vicinity of said recording temperature.
Since Ms shows a rapid decrease in the vicinity of the Curie point, a substance showing a relation Hk ~ 4~Ms at room temperature may show a relation Hk 2 4~Ms in the recording temperature range. As the component vertical to the substrate surface increases by the magnetization of the adjusting layer, said magnetization is further oriented vertical to the substrate surface by the exchange forces from the first and second magnetic layers. The exchange forces Heff(l 2) and Heff(2 3) 1~ working on the adjusting layer respectively from the first and second magnetic layers can be represented by:

Heff(l 2) = ~ /2M h Heff(2 3) = ~w23/2Ms2h2 wherein MS2 is the saturation magnetization of the adjusting layer, h2 is the thickness thereof, and ~w12 ~w23 are magnetic wall energies respectively between the first magnetic layer and the adjusting layer and between the second magnetic layer and the adjusting layer.
Therefore, in order to orient the magnetization of the adjusting layer in a direction vertical to the film surface in the recording temperature range by means of said exchange forces Heff(1-2) and Heff(2-3)' it is 1349~

1 advantageous to select a small saturation magnetization Ms and a thickness h2, as long as the easy direction of magnetization remains longitudinal at room temperature.
Example 10 A pregrooved and preformatted polycarbonate disk su~strate was placed in a sputtering apparatus with quaternary targets, and was rotated at a distance of 10 cm from the targets.
A Si protective layer of 500 A in thickness was sputtered from a first target, in argon gas, with a sputtering speed of 100 A/min., and a sputtering pressure of 5 x 10 3 Torr. Then a GdTbFe alloy was sputtered from a second target, in argon gas, with a sputtering speed of 100 A/min. and a sputtering pressure of 5 x 10 3 1~ Torr to obtain a first magnetic layer composed of Tbl2Gdl0Fe78 with a thickness of 300 A, TL = ca. 150~C
and HH = ca. 8 KOe. Fe was prevailing in the sub-lattice magnetization of the first magnetic layer.
Then a TbFeCo alloy was sputtered from a third target, in argon gas, with a sputtering speed of 100 A/
min. and a sputtering pressure of 5 x 10 Torr to obtain an adjusting layer, composed of Tb35Fe60Co5 having a thickness of 200 A and a coercive force of almost zero at the Curie temperature of ca. 170~C. The easy direction of magnetization of said adjusting layer was neither longitudinal nor vertical, and the external magnetic field required for orienting the magnetization in either 5b ~ 0-;~

1 direction was ca. 2.5 KOe.
Then a TbFeCo alloy was sputtered from a fourth target, in argon gas, with a sputtering speed of 100 A/
min and a sputtering pressure of 5 x 10 3 Torr to obtain S a second magnetic layer composed of Tb24Fe68Co8 with a thickness of 300 A, TH = ca. 180~C and HL = ca. 1.5 KOe.
Tb was prevailing in the sub-lattice magnetization of said second magnetic layer.
Subsequently a Si protective layer of 1000 A in thickness was formed by sputtering from the first target, in argon gas, with a sputtering speed of 100 A/min. and a sputtering pressure of 5 x 10 Torr.
After these layer formations, the above-mentioned substrate was adhered to a polycarbonate plate with hot-melt adhesive material to obtain a magnetooptical disk.
The effective bias magnetic field, caused by the exchange force on the second magnetic layer, was almost zero in thus prepared magnetooptical disk, when measured in the same manner as in Fig. 14.
Said magnetooptical disk was mounted on a record/
reproducing apparatus and was made to pass through, with a linear speed of ca. 8 m/sec., a unit generating a magnetic field of 2.5 KOe. The recording operation was then conducted with a laser beam of a wavelength of 830 nm, concentrated to ca. 1 ~m and modulated in two levels of 4 and 8 mW, with a duty ratio of 50% and a frequency 5'7 13 l~

1 of 2 MHz. The bias magnetic field was 100 Oe. Binary signals could be reproduced by irradiation with a laser beam of 1.5 mW.
The above-explained experiment was repeated with a magnetooptical disk already recorded over the entire surface. The previously recorded signal components were not detected in the reproduction, so that the possibility of overwriting was confirmed.
Example 11 Samples of the magnetooptical disk were prepared with the same process and materials as in the Example 10, except that the material and thickness of the adjusting layer were varied.
The prepared samples were subjected to the measurements of the effective bias field caused by the exchange force on the second magnetic layer and the recording characteristic in the same manner as in the Example 10. The results are summarized in Tab. 4.

5~
~ 3 ~
1 Tab. 4 . . f Effective bias Recording Ad ~ustlng Thlckness o Example . .fleld on 2nd character-layer ad~ustlng la.
layer lstlc Tb35Fe60Co5 200 A 0 Oe Good 11-1 None - 2000 Oe First-type record not possible 11-2 70 30 40 A 200 Oe Good 11-3 DyCo5 200 A 0 Oe Good 11-4 SmO 7FeO 3Fe~3150 A 300 Oe Good 10 11-5 Si 40 A 250 Oe First-type record sensitivity low, bit error rate high 11-6 Si 60 A 150 Oe First-type record not possible 1;~
The sample 11-1 wash not provided with the adjusting layer. In this case stable record bits could not be formed because the effective bias magnetic field, required for orienting the magnetization of the second magnetic layer in a direction stable with respect to that of the first magnetic layer, was larger than the coercive force of the second magnetic layer.
The Fe70Cr30 film employed in the sample 11-2 has a Curie point lower than 200~C and a longitudinal easy axis of magnetization. With a thickness not exceeding 30 A, the magnetization was oriented in the vertical direction by the exchange forces from the -irst 1 and second magnetic layers. Therefore the effective bias field on the second magnetic layer became larger and stable recording could not be achieved. On the other hand, with a thickness of 100 A or larger, vertical magnetization was not induced in the recording temperature range due to the excessively large saturation magnetiza-tion, so that the recording was not made due to the absence of the exchange force.
DyCo5 (magnetic transition point 50 - 80~C) and SmO 7Er0 3FeO3 (magnetic transition point ca. 110~C) employed in the samples 11-3 and 11-4 provided satisfac-tory recording in a thickness range of 100 - 400 A. This result is based on a change of the easy axis of magnetization from the longitudinal direction to the vertical direction in the recording temperature range (50 - 150~C), despite of the fact that the effective bias field on the second magnetic layer is almost zero at room temperature.
Si employed in the samples 11-5 and 11-6 is not magnetic. With a thickness of 40 - 60 A, the Si layer hinders the exchange coupling between the first and second magnetic layers, so that the measured effective bias field on the second magnetic layer was as small as 250 - 150 Oe. This effective bias field decreased with the rise of temperature, so that the first-type recording, in which the magnetization of the first magnetic layer is arranged in a direction sta~ler with respect to the 6 ~) 1 3 ~

1 magnetization of the second magnetic layer against the bias field was not possible at the recording temperature range.
Fig. 15 shows the effective bias fields on the s first or second magnetic layer in ordinate, as a function of temperature in abscissa, measured on the sample of the Example 10.
The first magnetic layer receives no effective bias field up to 80~C, but receives, from 90~C, a bias field for orienting the magnetization of said first magnetic layer in a stable direction with respect to the second magnetic layer. Said bias field monotonously decreased above 90~C to reach zero at the Curie point of the first magnetic layer. The second magnetic layer received no bias field over the entire temperature range measured. These results coincide with the satisfactory recording characteristic of the magnetooptical disk of the Example 10.
Then a sample was prepared by sputtering an adjusting layer of Tb35Fe60Co5, employed in the Example 10, with a thickness of 1000 A on a slide glass, and by forming a Si3N4 protective layer of 1000 A in thickness thereon. Fig. 16 shows the external magnetic field required to orient the magnetization of said Tb35Fe60Co5 layer into the vertical direction, in ordinate, as a function of temperature in abscissa, measured on the above-mentioned sample. The magnitude of said required ~1 13~00:~

1 external field decreases with the rise of temperature, and becomes about 500 Oe in the temperature range of 80~ - 90~C where the first magnetic layer starts to receive a large bias field. In the sample of the Example l0, it is assumed that the magnetization of the adjusting layer is oriented, with the rise of temperature, in the vertical direction at the interface between the first magnetic layer and the adjusting layer owing to the exchange force at said interface, and the magnetiza-tion of the adjusting layer in the vicinity of theinterface with the second magnetic layer is also oriented in the vertical direction in a temperature range of 80 -90~C, so that a large effective bias field emerges in this state between the first and second magnetic layers through the adjusting layer.
The above-mentioned adjusting layer, composed of a rare earth-transition metal alloy, can be optimized in composition, in consideration of the following factors:
(i) The rare earth-transition metal alloy shows a magnetic anisotropy in the vertical direction, when the rare earth element represents a range 12 - 28 atomic % in the rare earth and transition metal elements.
Outside said range, the easy axis of magnetization is in the longitudinal direction or in the direction of surface, possibly because of following two reasons. Firstly, the condition Hk 2 4~Ms for realizing a vertical magnetization cannot be satisfied, because of the large saturation .. ~. , . , , ~, .. . .. . . .

d.; ~

1 magnetization Ms, wherein Hk being the vertical anisotropic magnetic field. Secondly, the vertical magnetic anisotropy in a rare earth-transition metal alloy film is caused by the coupling of the rare earth 5 element and the transition metal element. The magnetic moment of the rare earth-transition metal element pair has a high probability of orientation in the vertical direction only in the above-mentioned percentage.
(ii) The rare earth-transition metal alloy employed in the adjusting layer, if rich in the rare earth element compared with the compensation composition, is increased in the rare earth content from a composition showing vertical magnetic anisotropy thereby increasing the saturation magnetization and facilitating the magnetization in the longitudinal direction. If the alloy is rich in the transition metal element compared with the compensation composition, the transition metal is increased further from the composition showing vertical magnetic anisotropy, thereby increasing the saturation magnetization and facilitating the magnetization in the longitudinal direction. If the curie point of the material is selected in the vicinity of the recording temperature, the saturation magnetization decreases in the vicinity of the recording temperature, thereby satisfying the condition Hk ' 4~Ms for vertical magneti-zation. In this manner it is possible to orient the easy axis of magnetization in the surface direction at ~s l~lUU~

I room temperature and in the vertical direction at the recording temperature.
The change of the easy axis of magnetization of the adjusting layer from the surface or longitudinal s direction to the vertical direction was experimentally confirmed in the following manner.
Three samples were prepared by sputtering, on glass substrates, magnetic layers of Fe, Tb5Gd5Feg0 or Tbl6Gdl6Fe68 of a thickness of 500 A as an adjusting layer, under an argon pressure of 5 x 10 Torr. On each sample a second magnetic layer of Tb24Fe74Co6 of a thickness of 500 A was formed without breaking the vacuum, and a Si3N4 protective layer of 700 A in thick-ness was formed thereon.
1~ Each sample was subjected to the measurement of the external magnetic field required to orient the magnetization of the adjusting layer into the vertical direction, as a function of temperature.
90 or Tbl6Gdl6Fe68 employed as the adjusting layer did not have a vertical easy axis of magnetization at room temperature.
Fig. 17 shows the external magnetic field required for vertical orientation in ordinate, as a function of the temperature in abscissa, obtained in said measurement.
The adjusting layer composed of Fe did not show orientation of the magnetization into the vertical 6~

1 direction, since the decrease of saturation magnetization is still small at 160~C. The Tb5Gd5Feg0 and Tbl6Gdl6Fe68, both being rare earth-transition metal alloy and having the easy direction of magnetization also in the surface direction, show a significant decrease in magnetization because the Curie point is in a range of 100 - 200~C.
Thus the magnetization can be oriented in the vertical direction with a limited external magnetic field when heated to about 100~C. Particularly Tbl6Gdl6Fe68, which is richer in the rare earth element compared to the compensation composition, shows easier orientation of the magnetization in the vertical direction with a smaller external field at a higher temperature, in comparison with Tb5Gd5Feg0 which is richer in the transition metal element.
Besides the required external field becomes smaller, in the temperature range of 70~ - 80~C, than the coercive force of the second magnetic layer. A
measurement for identifying whether the external magnetic field required for inverting the magnetization of the adjusting layer is dependent on the direction of magnet-ization of the second magnetic layer, namely whether an exchange force exists between the adjusting layer and the second magnetic layer, clarified that no exchange force was present at room temperature but a bias field caused by an exchange force of ca. 200 Oe was present at 90~ and 110~C.

; 8 1 Tbl6Gdl6Fe68, richer in the rare earth element than the compensation composition, shows an enhanced orientation of the easy direction of magnetization in the vertical direction at higher temperatures because 5 of the following two reasons.
Firstly, it is empirically known that, in an exchange-coupled combination of the adjusting layer and the second magnetic layer, a stronger exchange force is obtained in a combination in which both layers are rich in the rare earth element or in the transition metal than in a combination in which one layer is rich in the rare earth element while the other is rich in the transition metal. Thus the adjusting layer is earily oriented in the vertical direction, since the second la magnetic layer is a magnetic film vertically oriented to the film surface.
Secondly, the rare earth elements have lower Curie points in isolated state. Thus, in the rare earth-transition metal alloys, the rare earth element contributes more significantly to the decrease of magnetization at higher temperatures, if the composition is rich in the rare earth element. For this reason the compensation temperature is present above room temper-ature.

A composition not showing vertical easy direction of magnetization because of the excessively large magnetization of the rare earth element shows a decrease ' 13fl~ d .-i ~

1 of magnetization of said rare earth element at higher temperatures, so that the magnetization is represented by the magnetizations of the rare earth element and the transition metal which originally have vertical magnetic anisotropy.
Example 12 A pregrooved and preformatted polycarbonate disk substrate was placed in a sputtering apparatus with quaternary targets, and was rotated at a distance of 10 cm from the targets.
A Si3N4 protective layer of 700 A in thickness was formed by sputtering from a first target, in argon gas, with a sputtering speed of 100 A/min., and a sputtering pressure of 5 x 10 Torr~ Then a Tb~yFeCo alloy was sputtered from a second target, in argon gas, with a sputtering speed of 100 A/min. and a sputtering pressure of 5 x 10 3 Torr to obtain a first magnetic layer of Tbl5Dy5Fe76Co4 with a thickness of 300 A, TL =
ca. 150~C and HH = ca. 10 KOe. Fe and Co atoms were prevailing in the sub-lattice magnetization of the first magnetic layer. Then a TbGd Fe alloy was sputtered from a third target, in argon gas, with a sputtering speed of 100 A/min. and a sputtering pressure of 5 x 10 3 Torr to form an adjusting layer of Tbl6Gdl6 68 thickness of 200 A, and Curie point of ca. 160~C. Said adjusting layer did not show a vertical easy direction of magnetization at room temperature, and the external 1 magnetic field required for orienting the magnetization into the vertical direction was ca. 2 KOe at room temperature. Then a TbGdFeCo alloy was sputtered from a fourth target, in argon gas, with a sputtering speed of 100 A/min. and a sputtering pressure of 5 x 10 Torr to obtain a second magnetic layer of Tb20Gd5Fe67Co8 with a thickness of 300 A, TH = ca. 190~C and HL = ca.
1.8 KOe. Tb and Gd were prevailing in the sub-lattice magnetization of said second magnetic layer.
Subsequently a Si3N4 protective layer of 800 A
in thickness was formed by sputtering from the first target, in argon gas, with a sputtering speed of 100 A/
min. and a sputtering pressure of 5 x 10 Torr.
After these layer formations, the above-mentioned substrate was adhered to a polycarbonate plate with hot-melt adhesive material to obtain a magnetooptical disk.
Then there was made a measurement, with VSM
(vibrating magnetization measurer), of the external magnetic field inducing the inversion of magnetization of the first and second magnetic layers. The effective bias field cause by the exchange force on the second magnetic field was identified as almost zero.
Said magnetooptical disk was mounted on a record/
reproducing apparatus and was made to pass through, with a linear speed of ca. 8 m/sec., a unit generating a magnetic field of 2.5 KOe. The recording operation was ~ ~ 1 3 ~

I then conducted with a laser beam of a wavelength of 830 nm, concentrated to ca. 1 ~m and modulated in two levels of 4 and 8 mW, with a duty ratio of 50~ and a frequency of 2 MHz. The bias field at recording was 150 Oe.
Binary signals could reproduced by irradiation with a laser beam of 1.0 mW.
The above-explained experiment was repeated with a magnetooptical disk already recorded over the entire surface. The previously recorded signal components were not detected in the reproduction, so that the pos-sibility of overwriting was thus confirmed.
Example 13 Samples of the magnetooptical disk were prepared with the same process and materials as in the Example 12, except that the composition of the adjusting layer was modified.
Then there were measured the effective bias magnetic field caused by the exchange force on the second magnetic layer, and the recording characteristic, according to the same methods as in the Example 12. The results are summarized in Tab. 5.

ul o u l o c~
Tab. S

Easy direc- Coercive Effective Recording Example Adjusting layer Note tion of mag- force of on 2nd mag. character-netization adj. layer layer istic 13-1Tb Gd Fe 5 Fe rich Surface ~ - 0 Not good (lst type -~ 2.5 2.5 9 recordlng not possible) _ 1 ~
13-2Tb5Gd5FegO " " ~ - 0 Not good (lst type recordlng sensl-_ I tivity low) 13-3Tb Gd Fe " Vertical 0.3 KOe 0.5 KOe Not good (lst type 7 7 86 : ~ ' recording not possible) C~
13-4Tb9Gd9Fe82 " " 1.5 2.5 13-5_10 10 80 _ 2.5 ~~-~~~~ ~l -13 612 12 76 t ~ - 2.5 13-7Tbl3Gdl3Fe74 " " 1 2.0 13-8Tbl5Gdl5Fe70 " Surface - O Good ~_~
13-920 20 60 Good 13-1025 25 50 G~~dgood (1st type W
13-1140 40 20 0.1 recording sensi-tivity low) ~ 7~

1 In Tab. 5, the column "Note" indicates whether the composition of the adjusting layer is rich in Fe or in Tb and Gd.
Satisfactory recording characteristic was found in the samples 13-8 to 13-10, in which the rare earth elements (Tb, Gd) represented 20 - 50 atomic % in the adjusting layer.
All the samples were easy to magnetize in the surface direction, as these compositions were rich in the rare earth elements of Tb and Gd, compared to the compensation composition.
In the samples 13-1 to 13-3, the effective bias field on the second magnetic layer was almost zero because the adjusting layer was rich in the transition metal (Fe), but the recording of first type could not be achieved in stable manner, because the bias field caused by the exchange force between the first and second magnetic layers and acting to orient the magneti-zation of these layers in stable directions was small.
The samples 13-4 - 13-7, having a vertically magnetized adjusting layer, shows an effective bias field on the second magnetic layer in the order of 2.0 to 2.5 KOe, larger than the coercive force thereof.
The recording of first type could~be made in these samples, since the magnetization of the second magnetic layer is always oriented in a stable direction with respect to the magnetization of the first magnetic layer.

; 71 ~ 5 1 The sample 13-8 had the highest proportion of the rare earth element and a Curie point lower than 100~C. The recording of first type could be made satisfactorily because there was no sufficient bias 5 field caused by the exchange force between the first and second magnetic layers at the recording temperature.
Example 14 In the samples of the Example 13, the compositions of the first magnetic layer, adjusting layer and second magnetic layer were respectively rich in the transition metal, rich in rare earth element and rich in rare earth element.
In the present example, samples were prepared with same thicknesses, materials and structures as 1~ those in the Example 13, except that the first and second m~gnetic layers and the adjusting layer were selected from a transition metal-rich composition (TM) and a rare earth-rich composition (RE) of a same coercive force and a same Curie temperature. The samples were evaluated in the same manner as in the Example 13, as summarized in Tab. 6.

~ L, 134~q~

Tab. 6 Structure Effect. bias P1st mag. Adjust. 2nd mag. 1 Record ng layer layer layer 14-1 TM TM TM 1 KOe Not good (lst type recording not possible) 14-2 TM RE TM 0 Not good (lst type recording sensi-tivity low) 14-3 RE TM TM 0 "
14-4 RE RE TM 0 "
14-5 RE TM RE 0 Not good (lst type recording unstable) 14-6 RE RE RE 1 Not good (lst type recording not possible) 14-7 TM TM RE 0 Not good (lst type recording sensi-1~) tivity low) Tab. 7 CuYi~
Layer Type Composition Coerclve force point 1st mag. layer TM Tbl5Dy5Fe76Co4 10 KOe 15Q~C

RE Tbl6 5Dy5~5Fe74 4 10 150~C
Field for vertical orientation Adjust. layer TM 6 6 88 2 KOe 150~C
RE Tbl6Gdl6F 68 2 150~C
2nd mag. layer TM Tbl2Gd3Fe80c 5 L.8 190~C
RE Tb20Gd5Fe67Co8 1.8 190~C

7 ~ ~ ~3 ~ r~ G -~ ~

I In Tab. 6 there are shown, corresponding to the coercive force of the adjusting layer, the values of external magnetic field required for orienting the magnetization into the vertical direction. As will be i apparent from Tab. 7, the samples 14-1 to 14-4, with transition metal-rich second magnetic layer, showed a larger decrease in the coercive force at an elevated temperature than in the samples of the Example 13 with rare earth-rich composition, so that the coercive force of the second magnetic layer became insufficient for forming stable bits at the temperature of recording of first type. For this reason there was encountered a low sensitivity for the recording of first type or unstable recording.
lS The samples 14-3 to 14-6, with rare earth-rich first magnetic layers, showed a smaller decrease of the coercive force at-the elevated temperature, in comparison with the transition metal-rich compositions in the Example 13. Consequently the condition HH ~ HHeff ~ ~
for recording in the first magnetic layer at the estimated recording temperature of the first type. Thus the recording of first type could not be conducted, or the sensitivity therefor was low.
Also the samples 14-1, 14-2, 14-5 and 14-6, in ~hich the first and second magnetic layers were both rich in the rare earch element or in the transition metal, showed a strong exchange force even at a relatively 1~403 ~

1 low temperature when the medium was heated, and the recording of first type was only conducted unstable in all these samples, because the magnetization of the second magnetic layer was oriented in the stable direction with respect to that of the first magnetic layer.
The samples 14-1, 14-3, 14-5 and 14-7, with transition metal-rich adjusting layers, showed a smaller bias magnetic field caused by the exchange force acting on the first and second magnetic layers through the adjusting layer, in comparison with the samples of the Example 13 with rare earth-rich adjusting layers.
Consequently the sensitivity for the recording of first type was lowered.
From the foregoing results it can be concluded that the optimum combination of the compositions of the first and second magnetic layers and the adjusting layer corresponds to the case of the Example 13, in which the first magnetic layer, adjusting layer and second magnetic layer are respectively rich in transition metal, rare earth element and rare earth element.
In the foregoing examples, the recording film has triple-layered structure consisting of the first and second magnetic layers and the adjusting layer, but it is also possible to form, between the first magnetic layer and the substrate, a fourth magnetic layer of a strong magnetooptical effect in strong exchange-coupling with the first magnetic layer. In such case the sum of 1 3 l~i) 05~

1 the thicknesses of said first and fourth magnetic layers is preferably in excess of 200 A for increasing the reproduction output.
In the following there will be given more S detailed explanation on the structure of the magneto-optical recording medium of the present invention and variations thereof.
Thickness of magnetic layer In the medium shown in Fig. 3, the thickness 0 Ll of the first magnetic layer 2 and the thickness L2 of the second magnetic layer 3 are preferably determined in the following manner.
In the conventional exchange-coupled double-layered film, the magnetization of a record bit in either of two states is inverted in two magnetic layers according to the state to be recorded. On the other hand, in the present invention, the magnetization of the first magnetic layer 2, principally related to the reproduction, is inverted according to the state to be recorded, while the magnetization of the second magnetic layer 3, principally related to the recording, coincides with the direction of the field from the magnetic field generating unit 8 shown in Fig. 6 and remains unchanged.
Since the sensitivity of the magnetooptical recording medium is lowered as the thickness of the magnetic layer not contributing to the magnetooptical effect increases, it is necessary to minimize the total thickness Ll + L2 ~ 3 ~ 3 .~ ~

1 of said magnetic film and to optimize the combination of Ll and L2 so as to increase the magnetooptical effect at the reproduction. As will be explained in the following examples, the layer thicknesses should satisfy following conditions:
600 A ' Ll + L2 _ 1000 A
200 A '- Ll in order to obtain a satisfactory sensitivity and a large magnetooptical effect.
Example 15 A pregrooved and preformatted polycarbonate disk substrate was set in a sputtering apparatus with ternary targets, and was rotated at a distance of 10 cm from the targets.
A ZnS protective layer of 800 A in thickness was formed by sputtering from a first target, in argon gas, with a sputtering speed of 100 A/min. and a sputtering pressure of 5 x 10 Torr. Then a TbFe alloy was sputtered, in argon gas, from a second target with a sputtering speed of 100 A/min. and a sputtering pressure of 5 x 10 Torr to obtain a first magnetic layer of Tblg 5Fe80 5 with a thickness of 300 A, TL = ca. 140~C
and HH = ca. 8 KOe.
Then a TbFeCo alloy was sputtered, in argon gas, with a sputtering pressure of 5 x 10 3 Torr to obtain a second magnetic layer of Tb24 5Fe68Co7 5 thickness of 400 A, TH = ca. 190~C and HL = ca. 0.8 KOe.

1 3 ~ 8 1 Subsequently a ZnS protective layer of 3000 A
in thickness was formed by sputtering, in argon gas, from the first target with a sputtering speed of 100 A/
min. and a sputtering pressure of 5 x 10 Torr.
After these layer formations, the above-mentioned substrate was adhered to a polycarbonate plate with hot-melt adhesive material to obtain a magnetooptical disk.

O O O O O
(600 A < Ll + L2 = 700 A < 1000 A, 200 A < Ll = 300 A) Said magnetooptical disk was set on a record/
reproducing apparatus, and was made to pass through, with a linear speed of ca. 8 m/sec., a unit for generating a magnetic field of 2.5 KOe. The recording operation was then conducted with a laser beam of a wavelength of 830 nm, concentrated to ca. 1 ~m and modulated in two 1~ levels of 4 and 8 mW, with a duty ratio of 50% and a frequency of 2 MHz. The bias field was 100 Oe.
Binary signals could be reproduced by irradiation with a laser beam of 1.5 mW.
The above-explained experiment was repeated on a magnetooptical disk already recorded over the entire surface. The previously recorded signal components were not detected in the reproduction, and the possibility of overwriting was thus confirmed.
Example 16 Samples of magnetooptical disk were prepared with the same process and materials as those in the Example 15, except that the thicknesses of the first 7~
~ 3 ~0.;8 l and second magnetic layers were modified.
For evaluating the reproduction signal from each sample, the reflectance at 830 nm and the Kerr rotation angle were measured. Also there was calculated 5 the product of aquare root of reflectance and Kerr rotation angle, which is usually accepted as an index for the reproducing performance in case the photosensor is composed of a photodiode. These results are summarized in Tab. 8.

Tab. 8 ~ ~9 50.8 Sl4 sl.3~-magn ~ oopti~al e ~ ect(Kerr ro,ation angle)(') 600 0.195 0.186 O.lnl 0.118 0.178~-ref1~ ctance 19.4 21 \ 21.fi 21.7 21.C~r~~tance x Kerr rot~tion ~ngle ~t 550. \ ~\
\
38.2 ' 45.6 50.2 52.~ 53.1 53.0 500 0.207 ' 0.195 0.185 0.179 0.-117 0.176 17.4 '~ 20.1 21.6 22.2 22.3 22.2 ~50 ~ \
33 1 41-7 ' 48.2 52.4 54.5 55.3 400 0.211 0.20~ ' 0.191 0.182 0.176 0.173 15.2 18.8 '~21.1 22.3 22.9 23.0 ~ , :, '~ \ ~ .
350 '' 29.7 37.8 44.4 \ 49.7 53.3 55.3 ;~
~A)300 0.191 o.ao2 0.196 0.186 0.177 0.171 13.0 17.0 19.7 21.4 22.4 22.9 250 ' ~
27.0 33.7 37.8 41.8 " 45.4 48.2 49.9 200 0.13~ 0.177 0.193 0.190 '0.181 0.172 0.167 9.87 14.2 16.6 18.2 19.3 '~_~ 20.0 20.4 24.0 3Q.9 32.8 35.0 37.7 ~0.1 41.9 42.9 150 0.093 0.151 0.184 0.191 0.186 0.177 0.170 0.185 ~A
7.31 12.0 1~.0 15.3 16.2 16.9 17.3 17.4 15.6 26.4 25.9 26.0 27.1 28.C 30.0 31.0 100 0.051 0.117 0.168 O.lgl 0.193 0.186 0.178 0.172 '~
3.52 9.0~ 10.6 11.3 11.9 12;3 12.7 12.8 C
13.0 18.3 IS.8 1~ .1 1'4.fi 15.2 15.7 i~x~
S0 0.019 0.077 0.145 0.188 0.203 0.201 0.193 ~.186 1.81 5.0~ 6.03 6.2~ 6.37 6.52 6.fi7 6.79 0 50 100 150200 250 300 350 ~00 ~S0 500 SS0 6~ fiS0700 (L2) (A ~

&~ 0 ;~ ~

1 In the record bit of the present invention, the magnetization of the second magnetic layer is not changed by the state of record and does not contribute to the magnetooptical effect. In case the first magnetic layer 2 is thin, the intensity of reflected light is reduced and the aforementioned performance index is also reduced if the second magnetic layer 3 is thin or absent.
The intensity of the reflected light becomes almost constant if the total thickness Ll + L2 of the first and second magnetic layers 2, 3 exceeds 400 A.
It is also found out that said performance index and Kerr rotation angle become saturated at substantially constant values when the total thickness Ll + L2 of the first and second magnetic layers is equal to or larger than 600 A, but, if the thickness Ll of the first magnetic layer 2 is less than 200 A, the Kerr rotation angle is saturated at a smaller value than the saturation value when said thickness Ll is larger than 200 A even when the thickness L2 of the second magnetic layer 3 is increased. From these results it is concluded that the conditions 600 A < Ll + L2 and 200 A < Ll are preferable.
These samples were subjected to a test of recording and reproduction in the same manner as in the Example 15, and satisfactory reproduction signal as in the Example 15 was obtained from the samples in which Ll + L2 was equal to or larger than 600 A.
On the other hand, when Ll + L2 was equal to 8 1 1 3 i~

1 1000 A, the laser powers required for the recordings of first and second types were respectively 2.5 times of those required when L1 + L2 was 600 A, and said powers rapidly increased when L1 + L2 exceeded 1000 A. Thus a ¢ondition Ll + L2 < 1000 A was concluded preferable.
As explained in the foregoing, though the second magnetic layer 3 does not change its magnetization during the reproduction of the recorded signals of first and second types, thus not contributing to the magneto-optical effect, the presence of said second magneticlayer provides a Kerr rotation angle and a reproduction performance index equivalent to the case where the first magnetic layer 2 has a thickness Ll + L2.
For the purpose of comparison, the reflectance, Kerr rotation angle and performance index were determined by applying a field stronger than the coercive force of the first magnetic layer in each sample, thereby inverting the magnetization of the second magnetic layer 3 as well as that of the first magnetic layer both in the recordings Of first and second types, just as in the conventional exchange-coupled double-layered magnetic film. The results are summarized in Tab. 9.

- Ta-,. 9 43.9 47.~.6~-- 50.2 50.0 49.5 49.1 600 0.195 0.180.181~- 0.178 0.178 0.178 0.178 19.~ 2~. 21.~ 21.2 2l.l 2~.g 20.7 " \ _ magnetooptical er ect(K-rr rotation angle)(') 550 ~ \ refle~tance refle~tance x Ker- rotat ion ar gle . 38.2 43.5 \ 47.6 ~9.9 50.C 50.3 50.1 ~9.7 500 0.207 0-19s 0.185 0.179 0.177 0.176 0.176 0.176 ~ 17.4 19.2 20.5 21.1 21.3 21.1 21.0 20.8 ; ~ 450 ~ \
- 33.1- 38.6 ~4.3 \ 48.7 5r.2 52.2 52.2 51;9 ' 400 0.211 0.203 0.191 0.1820.176 0.173 0.172 0.172 15.2 17.4 19.4 20.8 21.5 21.7 21.7 21.5 350 '~
29.7 34.7 40.5 ' 46.4 51.2 54.1 55.3 55.5 300 0.191 0.202 0.197 ~ 0.~0.177 0.172 0.16g ~.167 (A) 13.0 15.6 18.0 ~ 20.1 21.5 22.4 22.7 22.7 27.0 33.1 38.0 ~43.7\ 49.7 54.7 58.0 59.5 200 0.13~ 0.178 0.19~ ~0.192 0.182 0.174 0.168 0.165 9.87 14.0 16.7 19.1 21.2 22.8 23.8 24.2 24.0 33.1 37.7 42.5 48.3 5:.9 58.2 60.8 150 0.09~ 0.153 0.186 0.193 ~ 0.187 0. 79 0.171 0.167 7.3 12.9 16.2 18.7 ~ 20.9 2--.8 24.1 24.8 15.6 33.4 38.1 41.7 ~ ~6.6 52.2 57.2 60.7 100 0.051 0.12 0.170 0.193 ' 0.195 0.187 O.l79 0.173 3.S2 11.5 15.7 18.3 '~ 20.5 22.6 24.2 25.3 13.0 33.3 39.5 41.5 ~ 4~.7 49.~ 54.6 ~ 58.9 50 0.019 0.078 0.147 0.190 \ 0.207 0.202 0.194 0.187 1.81 9.32 15.1 18.1 ~20.2 22.2 24.1 25.5 0 50 100 150 240 250 ~ 350 ~ ~0 500 S50 600 6~ 7 Second magnetic laye~L2)(A) 8~ ?

1 In this case, large rotation angle and index can be obtained if the thickness of the second magnetic layer is in a range of 300 - 400 A, even if the first magnetic layer 2 is thinner than 200 A, because the 5 second magnetic layer 3 shows a magnetooptical effect induced by the inversion of magnetization, and because the second magnetic layer 3 has a higher Curie point and shows a larger Kerr rotation angle than the first magnetic layer 2 at a same thickness. However the contribution of the second magnetic layer 3 in the magnetooptical effect is negligible when the thickness of the first magnetic layer 2 exceeds 200 - 250 A.
These results coincide with those of Tab. 8 that the presence of the second magnetic layer 3 provides a Kerr rotation angle and a performance index equivalent to those when the thickness of the first magnetic layer 2 is equal to Ll + L2.
Also when the ZnS protective layers in the samples of the Example 15 are replaced by those of Si3N4, SiC, SiO or A12O3, the relationship of the thicknesses of the first and second magnetic layers providing saturated rotation angle and performance index remains unchanged.
SiC, having a refractive index larger than that of Si3N4 or ZnS, increased the index by about 10%, but the recording sensitivity was lowered for a same thickness Ll + L2. On the other hand, SiO or A12O3, having refractive indexes lower than those of Si3N4 or ZnS, 8~ 13~00 )~

1 decreased the performance index by about 10%.
Also the relationship of the thicknesses of the first and second magnetic layers providing saturated rotation angle and performance index was not affected 5 by a change of the material constituting the first magnetic layer 2, from TbFe to GdTbFe, TdFeCo, GdTbFeCo or DyTbFeCo.
Compensation temperature of magnetic layers In the magnetooptical recording medium as shown in Fig. 3, the compensation temperature THComp of the second magnetic layer 3 is preferably positioned between room temperature and Curie temperature TH of said second magnetic layer. Also such medium allows a recording process utilizing said compensation temperature. This ~~ process will be explained in the following, separately L THComp and a case of TH < T
(i) Case TL ~ THcomp Referring to Fig. 5, in one of the states 4b prior to the first prelim;nAry recording, in which the magnetizations of the first and second magnetic layers are in a same direction, said magnetic layers are stabilized by the exchange force, so that the magnetiza-tion of the first magnetic layer can only be inverted by a magnetic field HH' + HHeff', which are respectively the coercive force of the first magnetic layer and the exchange force acting on said layer at a temperature t and which are both functions of the temperature t, while 8 l) 1 3 ~

1 the magnetization of the second magnetic layer can only be inverted by a magnetic field HL' + HLeff', which are respectively the coercive force of the second magnetic layer and the exchange force acting on said layer at a 5 temperature t and which are both functions of the temperature t. For this reason the preliminary recording bit 4c can be stably formed, even in the presence of a certain bias field in either direction. However, in the other of the states 4b shown in Fig. 5, in which the magnetizations of the first and second magnetic layers are mutually oppositely directed, each magnetic layer receives an exchange force to invert its magnetization.
Thus, there will result a situation (a) or (b) explained in the following.

(a) If a situation HH' - HHeff' < 0 occurs when the magnetic layers are heated close to TL at the preliminary recording of first type, the magnetization of the first magnetic layer is inverted and arranged in a stable direction with respect to the magnetization of the second magnetic layer, thus completing the prelimin~ry recording of first type.
(b) On the other hand, if a condition HL' -HLeff' < 0 is reached before the inversion of magnetiza-tion of the first magnetic layer, the magnetization of the second magnetic layer is inverted and arranged in a stable direction with respect to the magnetization of the first magnetic layer. Therefore, the preliminary 8b L3 100 ~

1 recording of first type, requiring the inversion of magnetization of the first magnetic layer, becomes impossible.
Therefore the value of HL' should be increased as far as possible, and, for this purpose, HL should be maximized, but can only be increased to ca. 2 KOe because the second magnetic layer needs to be uniformly magnetized by the magnetic field generating unit 8 shown in Fig. 6.
Consequently, in order to maximize the coercive force of the second magnetic layer in the vicinity of the temper-ature TL, it is important to select the compensation temperature THcOmp of the second magnetic layer close to TL.
In this manner it is rendered possible to prevent the decrease in the coercive force HL' of the second magnetic layer in the course of heating toward the compensation temperature THComp, because the coercive force of the second magnetic layer becomes infinite at the compensation temperature THComp even if the satura-tion magnetization Ms of the second magnetic layer isdecreased.
As will be apparent from the following examples, it is possible to achieve the preliminary recording of the first type in stable manner, by selecting a second magnetic layer satisfying conditions:

L Hcomp;
(coercive force of second layer at temperature 8 ~ L ~ 5 ~

1 t)/(coercive force of second layer at room temperature (= HH)) > 0.5.
Among the aforementioned substances, the above-mentioned conditions can be satisfied by a TbGdCo alloy richer in Tb than the compensation composition. Also there may be employed substances obtaining by adding suitable impurities such as another rare earth element or another transition metal to the above-mentioned ternary compound.
Example 17 A pregrooved and preformatted polycarbonate disk substrate was set in a sputtering apparatus with ternary targets, and was rotated at a distance of 10 cm from the targets.

After the sputtering apparatus was evacuated to 1 x 10 6 Torr or lower, a ZnS protective layer of 1000 A in thickness was formed by sputtering, in argon gas, from a first target with a sputtering speed of 100 A/min.
and a sputtering pressure of 5 x 10 Torr. Then a TbFe alloy was sputtered, in argon gas, from a second target with a sputtering speed of 100 A/min. and a sputtering pressure of 5 x 10 Torr to obtain a first magnetic layer of Tbl8Fe82, with a thickness of 300 A, TL = ca.
130~C and HH = ca. 10 KOe.

Then a TbFeCo alloy was sputtered, in argon gas, with a sputtering pressure of 5 x 10 3 Torr to obtain a second magnetic layer of Tb27Fe64Cog with a thickness 8~ ~3'~0~ ?~

I of 500 A, TH = ca. 190~C, THComp = 210~C and HL = ca. l KOe.
Subsequently a ZnS protective layer of 3000 A
in thickness was formed by sputtering, in argon gas, from the first target with a sputtering speed of A and a sputtering pressure of 5 x 10 3 Torr.
After these layer formations, the above-mentioned substrate was adhered with a polycarbonate plate with hot-melt adhesive material to obtain a magnetooptical disk.
The second magnetic layer showed a coercive force of ca. 700 Oe around the Curie point (130~C) of the first magnetic layer, when méasured by the method explained in the Example 18.

This magnetooptical disk was mounted on a record/
reproducing apparatus, and was made to pass through, with a linear speed of ca. 8 m/sec., a unit generating a magnetic field of 2.5 KOe. The recording was conducted with a laser beam of a wavelength of 830 nm, concentrated to ca. 1 ~m and modulated in two levels of 4 and 8 mW, with a duty ratio of 50% and a frequency of 2 MHz. The bias field was 100 Oe in a direction to invert the magnetization of the second magnetic layer. Binary signals could be reproduced by irradiation with a laser beam of 1.5 mW.
The above-explained experiment was repeated with a magnetooptical disk already recorded over the 8 ~ O ;~i ,3 1 entire surface. The previously recorded signal components were not detected in the reproduction, and this result confirms that the overwriting is possible.
Example 18 According to a process similar to that in the Example 17, there were prepared a magnetooptical disk sample 18-1 having a ZnS layer of 1000 A, a first magnetic layer of Tbl8Fe82 of 300 A, a second magnetic layer of Tb27Fe64Cog of 500 A, and a ZnS layer 1000 A formed in this order on a glass substrate, and a sample 18-2 of the same structure except that the second magnetic layer was replaced by Tbl5Fe81Co4.
B-H loops were determined by measuring the magnetic inducing the inversion of magnetization, by means of magnetic Kerr effect, in the samples 18-1 and 18-2, at various temperatures. In this manner obtained were the temperature characteristic of the coercive force (Fig. 18) and the temperature dependence of the differ-ence of coercive force and exchange force (Fig. 19), for the Tb18Fe82 layer (hereinafter called layer A), Tb27Fe64Cog layer (layer B) and Tbl5Fe81Co4 (layer C).
As shown in Fig. 18, the layer A showed, at room temperature, a coercive force HH of ca. 10 KOe and a Curie point Tl of ca. 130~C.

Also at room temperature, the layers B and C
showed a coercive force HL of ca. 1 KOe and a curie point TH of ca. 190~C.

91) ~L3 'I ~ O ~
1 The compensation temperature was ca. 210~C and ca. -130~C respectively for the layers B and C. The coercive forces of the layers B and C around TL were respectively ca. 70% and ca. 30% of those at room S temperature.
In Fig. 19 showing the temperature dependence of the difference of coercive force and exchange force, broken lines indicate HL' - HLeff' for the layer B or C, while solid lines indicate HH' - HHeff' for the first magnetic layer A.
The lines marked as A/B are related to the sample 18-1, while those marked as A/C are related to the sample 18-2.
In practice, only the smaller one of HH' - HHeff' lS and HL' - HLeff' is measurable. The larger one cannot be measured because, when the external magnetic field is increased, the magnetization of a magnetic layer with a smaller coercivity is inverted at a point lower than the inverting magnetic field.

The meaning of Gi.l9 will be explained in the following.
When the inverting magnetic field is measured with increasing temperature, HL' - HLeff' of the second magnetic layer initially has a positive value of 200 -3000 Oe. However, at a temperature lower than TL by 20~ - 50~C, HH' - HHeff' of the first magnetic layer becomes negative, so that the second magnetic layer is 9 ~ 0 0 ,~ ~

1 spontaneously oriented in the stable direction with respect to the magnetization of the first magnetic layer.
At a temperature higher than TL, the first magnetic layer is no longer magnetized to annulate the exchange s force, so that only the coercive force of the second magnetic layer is measured.
While the sample 18-1 starts the prelim;n~ry recording of the first type from a temperature lower than TL by about 50~C and shows the inversion of magnet-ization of the first magnetic layer due to a strongexchange force, the sample 18-2 starts said preliminary recording at a temperature lower than TL by about 20~C
due to a weaker exchange force.
This difference results from a difference in the temperature-dependent coercive force of the second magnetic layer, or in the compensation temperature. The value of ~coercive force - exchange force) of the first magnetic layer changes from a large positive value to a negative value as the sample temperature is increased.
Therefore, if the value of (coercive force - exchange force) of the second magnetic value is a relatively large positive value at the temperature range of prelim-inary recording of the first type, the first magnetic layer receives a strong exchange force (value of (coercive force - exchange force) being a large negative value) from a relatively low temperature at the preliminary recording of the first type, thereby orienting the 9~ ~floo~

1 magnetization into a stable direction with respect to the magnetization of the second magnetic layer.
In conclusion, it is concluded that the temper-ature dependence of the coercive force of the second magnetic layer and the magnitude relationship between the compensation temperature of the second magnetic layer and the Curie temperature of the first magnetic layer are the factors determining the stability and sensitivity of the preliminary recording of first type.
Example 19 Samples 19-1 to 19-6 of magnetooptical disk were prepared with the same process, structure and thicknesses as in the Example 17, except that the substance of the second magnetic layer was changed, and subjected to the evaluation of recording and reproduction in the same conditions as in the Example 17. The results are shown in Tab. 10.

~n o c,- o ~il ~

Tab. 10 , . Compensation Ratio of Threshold Evaluation Evaluation Composltlon of T of 2nd H temp. of 2nd coercive force value for of first of second Example 2nd mag. layer mag. layer (atom. ~) (~C) mag. layer at TL to room fir¢t type type record type record Hcomp temp. recording at 4 mW at 8 mW

27 64 9 190 - 210 0.7 3.5 mW + +
15 81 4 190 below room 0.3 4.5 mW - +
temp.
19-2 Tbl5Fe71.5 13.5 240 " 0.5 4 mW + t CD

15 66 19 280 3.5 mW +
27 67 6 170 240 0.7 3,5 mW +
- 27 61 12 215 210 0.8 3.5 mW +
19-6 Gdl3Tbl4Fe64Cog 197 220 0.8 3.5 mW + +

.

C

9~ ~3'~0~;~

1 However the coercive force HL of the second magnetic layer is about 1 KOe in all the samples. In Tab. 10, "Ratio of coercive force at TL to room temp."
indicates the coercive force at the temperature TL ~f a the first magnetic layer (ca. 130~C in the present exampleJ divided by the coercive force at room temper-ature.
The threshold value for the recording of first type indicates the laser power enabling the recording.
The recording of first or second type was evaluated as "+" if a satisfactory reproduction signal of a C/N ratio of about 40 dB could be obtained, as "+" if the recorded signal could be confirmed, and as "-" if recording was not made.
As shown in Tab. 10, the recording of first type could be satisfactorily made only in the samples having the ratio of coercive force at TL to that at room temperature is equal to or larger than 0.5 (samples 17, 19-2 to 19-6), in which the compensation temperature of the second magnetic layer is higher than the Curie temperature of the first magnetic layer, or the Curie temperature TH of the second magnetic layer is higher than that in other samples (case of samples 19-2 and 19-3)-Similarly the recording of second type could be satisfactorily made in the samples having the Curie temperature TH of the second magnetic layer lower than 9~ 13~ 05~

l 200~C (samples 17, 19-4, 19-5 and 19-6).

L Hcomp p ined above can be utilized in a recording process as shown in Fig.
20, in which there are shown a first magnetic layer 2 and a second magnetic layer 3. Various states of magnetization of said layers are represented by 44a - 44g.
In the recording process, at a position different from the recording head, there is applied a downward external magnetic field HE of a magnitude enough for magnetizing the second magnetic layer of a coercive force HL in one direction but insufficient for inverting the magnetiza-tion of the first magnetic layer of a coercive force HH, and a downward bias magnetic field HB for facilitating the recording in the second magnetic layer is applied at the position of the recording head.
Prior to the explanation of the steps of said recording process, there will be briefly explained the states 44a - 44g and the transitions therebetween.
44a and 44g indicate two different record states 20 at room temperature. Heating with a laser beam causes a transition in the order of 44b, 44c and 44d. 44b and 44f, or 44c and 44e represent different states at a substantially same temperature. An arrow +~ indicates a reversible magnetizing process, while arrows ~ and indicate irreversible magnetizing processes. The compensation temperature of the second magnetic layer is positioned between 44b and 44c, or between 44c and 44f.

9b ~ g)~
,.

1 Fig. 20 shows a case in which the rare earth lattice magnetization is prevailing in the first and second magnetic layers. In such case, the state 44g in which the magnetizations of the layers are mutually in a same direction is stable, while the state 44a in which the magnetizations are mutually oppositely directed is unstable, due to the mutual exchange effect of the two layers, and an interfacial magnetic wall exists in said unstable state 44a. However the coercive force of the second magnetic layer has to be so adjusted as that said unstable state can be maintained even in the absence of the magnetic field. In the room temperature state (44a, 44g), the magnetization of the second magnetic layer with smaller coercive force is always oriented downwards by the external magnetic field HE.
In the following there will be explained the steps of the recording process.
When the temperature is raised from the state 44a, the coercive force of the first magnetic layer decreases while that of the second magnetic layer increases as shown in Fig. 22. As the magnetizations of two layers tend to be oriented in a same direction by the exchange force, the magnetization of the first magnetic layer is inverted downwards as shown by 44b.
If the temperature is lowered from this state, the magnetization remains unchanged to reach the state 44g.

If Lhc ~cmpQraturo ~ 6 r~ from tho st~te ~ to th~ t ,4~re~ ~ f ~e ~ a~ f~o~ f~ 44 c~J'e~ r 4~ ,~ f~ ~ 4 ~,~ ~zO~ e~ o of ~ ,*4~ ~ ~4 ~~ ~ f~a~ ~ ~,c~ 7~

134giOa~
9~

1 If the temperature is further raised from the state 44b to a state 44c across the compensation temperature TcOmp of the second magnetic layer, the magnetization thereof is reversibly inverted. If the s temperature is further elevated, the coercive force of the second magnetic layer is reduced, so that the magnetization thereof is inverted by the bias field HB as shown by 44d. If the temperature is lowered from this state, the magnetization remains unchanged, and the magnetization of the second magnetic layer is reversibly inverted when crossing the compensation temperature TComp. At the same time the first magnetic layer starts to show upward magnetization due to the exchange force. The coercive force of the second 1~ magnetic layer becomes smaller when it is cooled to room temperature, and the magnetization thereof is inverted by the external field HE. However, the first magnetic layer, because of the large coercive force thereof, does not show inversion of the magnetization by the external field HE but retains the record state. In this manner, by the application of a laser power corresponding to the temperature of 44d, both states 44a and 44g are transformed to the state 44a.
Thus different states of magnetization can be obtained by the application of different laser powers, and this is a principle of overwriting.
Fig. 21 shows a case in which transition metal 1310~
9~

1 sub-lattice magentization is prevailing in the first magnetic layer, while rare earth sub-lattice magnet-ization is prevailing in the second magnetic layer.
In such case a state 45a, in which the magentizations of both layers are mutually oppositely directed, is a stable state, and a state 45g, in which the magnetizations of both layers are mutually in a same direction, is an unstable state, due to the exchange effect between the two layers, and an interfacial magnetic wall exists in said unstable state 45g. In a similar manner as shown in Fig. 20, the application of a laser power corresponding to the temperature of , a state 45b brings the states 45a and 45g to the state 45a, while the application of a laser power corresponding to the temperature of a state 45d brings the states 45a and 45g to the state 45g. Consequently, different states of magnetization can be again obtained by the application of different laser powers. In this manner overwriting can be achieved.

The magnitude of the field HB is smaller than that of the field HE. If HB is larger than HE, the field HE becomes unnecessary but the control of the compositions and layer thicknesses of the medium becomes more difficult. More specifically, if HB ~ HE~
HB has to be considerably large, and it is considerably difficult to realize the state 44f in Fig. 20 or 45b in Fig. 2l by the exchange force, against such large ~3.l 1 magnetic field. Since the second magnetic layer has a compensation temperature as explained before, the magnetization of the second magnetic layer in the state 44f in Fig. 20 or 45f in Fig. 21 is stabilized against the magnetic vield HB.

Example 20 A pregrooved and preformatted polycarbonate disk substrate was set in a sputtering apparatus with ternary targets, and was rotated at a distance of 10 cm from the targets.
A Si3N4 protective layer of 700 A in thickness was formed by sputtering, in argon gas, from a first target with a sputtering speed of 100 A/min. and a sputtering pressure of 5 x 10 Torr. Then a TbFe alloy was sputtered, in argon gas, from a second target with a sputtering speed of 100 A/min. and a sputtering pressure of 5 x 10 Torr to obtain a first magnetic layer with prevailing Tb sub-lattic magnet-ization, with a thickness of 500 A, TL = ca. 130~C andHH = ca. 5 KOe.
Then a Gd-Tb-Fe-Co alloy was sputtered, in argon gas, from a third target with a sputtering speed of 100 A/min. and a sputtering pressure of 5 x 10 Torr to obtain a second magnetic layer with prevailing GdTb sub-lattice magnetization, with a thickness of 800 A, TH = ca. 220~C, HL = ca. 1.5 KOe and TComp = ca.

140~C.

lOU 131!39.:~

1 o Subsequently a Si3N4 protective layer of 700 A in thickness was formed by sputtering, in argon gas, from the first target with a sputtering speed of 100 A/min. and a sputtering pressure of 5 x 10 Torr.
After these layer formations, the above-mentioned substrate was adhered to a polycarbonate plate with hot-melt adhesive material to obtain a magneto-optical disk. Said magnetooptical disk was set on a record/reproducing apparatus, and the recording was conducted with a laser beam of a wavelength of 830 n~, focused to ca. 1.5 ~m and modulated in two levels of 2.7 and 5.5 mW, with a duty ratio of 50 % and a frequency of 2 MHz, and with a linear speed of ca. 8 m/sec and under the application of a bias magnetic field of 200 Oe and an external magnetic field of 2 KOe.
Subsequently binary signals could be reproduced by irradiation with a laser beam of 1 mW.
After the above-explained experiment, recording was conducted on the same track with a same power and a frequency of 3 MHz. The previously recorded signals were not detected, and the possibility of over-writing was thus confirmed.
Example 21 A magnetooptical disk was prepared in the same manner as in the Example 20, except that, for the first magnetic layer, a TbFe alloy was sputtered from the second target with a sputtering speed of 100 A/min.

lol 1~'IOO~3 1 and a sputtering pressure of 5 x 10 Torr to obtain a magnetic layer with prevailing Fe sub-lattice magnet-ization, with a thickness of 500 A, TL = ca. 125~C, and HH = ca. 4 KOe, and said disk was evaluated in the 5 same manner.
The evaluation clarified that the overwriting was possible also.

(ii) Case THcOmp ~ L
The recording process is conducted in the following manner:
(a) To the recording medium there is applied, at a position different from the recording head, a magnetic field B enough for magnetizing the second 1~ mangetic layer of a coercive force HL in a direction but insufficient for inverting the magnetization of the first magnetic layer of a coercive force HH.
(b) Then there is conducted, according to the input signal, either the preliminary recording of first type in which a bias magnetic field is applied by the recording head and a laser power enough for heating the medium close to the compensation temperature THComp is simultaneously applied, thereby orienting the magnet-zation of the first magnetic layer in a stable direction with respect to the magnetization of the second magnetic layer without varying said magnetization, or the preliminary recording of second type in which a laser 13~a~

1 power enough for heating the medium close to the higher of the Curie temperatures TL, TH ~f the first and second magnetic layers is applied simultaneously with the application of the bias magnetic field, thereby S inverting the magnetization of the second magnetic layer and simultaneously magnetizing the first magnetic layer in a stable direction with respect to the magnetization of said second magnetic layer.
(c) Then the medium is so moved that the pre-liminarily recorded bit passes through the afore-mentioned magnetic field B, whereby the bit formed by the preliminary recording of first type does not change the direction of magnetization in the first and second magnetic layers, while the bit formed by the preliminary 1S recording of second type does not change the direction of magnetization of the first magnetic layer but inverts the magnetization of the second magnetic layer in a direction same as that of said magnetic field B.
In this case the Curie temperatures TL, TH ~f the first and second magnetic layers should satisfy a relation TL < TH. On the other hand, the condition for the recording of first type is HH' - HHeff' < 0, and HH' should preferably decrease rapidly at elevated temperature. Consequently, the compensation temperature TcOmp of the first magnetic laysr should preferably be lower than room temperature. The relation of the compensation temperature, Curie temperature and ~ , ~

lo~ 13~d j~

1 coercive in these magnetic layers is shown in Fig. 23.

Example 22 A pregrooved and preformatted polycarbonate 5 disk substrate was set in a sputtering apparatus with ternary targets, and was rotated at a distance of 10 cm from the targets.
A ZnS protective layer of 1000 A in thickness was formed by sputtering, in argon gas, from a first target with a sputtering speed of 100 A/min. and a sputtering pressure of 5 x 10 Torr. Then a TbFe alloy was sputtered, in argon gas, from a second target with a sputtering speed of 100 A/min. and a sputtering pressure of 5 x 10 3 Torr to obtain a first magnetic layer of Tbl8Fe82 with a thickness of 300 A, TL = ca.
130~C, a compensation temperature below room temperature, and HH = ca. 10 KOe.
Then a TbFeCo alloy was sputtered, in argon gas, witha sputtering pressure of 5 x 10 Torr to obtain a second magnetic layer of Tb27Fe51CogCul3 with a thick-ness of 500 A, Th = ca. 210~C, HL = ca. 1 KOe and a compensation temperature of 100 ~C.
Subsequently a ZnS protective layer of 1000 A
in thickness was formed by sputtering, in argon gas, from the first target with a sputtering speed of 100 A/min. and a sputtering pressure of 5 x 10 Torr.
After these layer formations, the 10~ 13~ ).?8 1 above-mentioned substrate was adhered to a polycarbonate plate with hot-melt adhesive material to obtain a magnetooptical disk.
Thus prepared was mounted on a record/reproducing apparatus, and was made to pass through a unit generating a magnetic field of 2.5 KOe with a linear speed of ca.
8 m/sec. The recording was then conducted with a laser beam of a wavelength of 830 nm, concentrated to ca. 1 ~m and modulated in two levels of 4 and 8 mW with a duty ratio of 50 % and a frequency of 2 MHz. The bias field was 100 Oe in a direction to invert the magnetization of the second magnetic layer. Binary signals could then be reproduced by irradiation with a laser beam of 1.5 mW.
1~ The above-explained experiment was repeated with a sample disk recorded over the entire surface.
The previously recorded signal components were not detected in the reproduction, and the possibility of over-writing was confirmed in this manner.

Example 23 A sample of magnetooptical disk was prepared in the same manner as in the Example 22, except that the first magnetic layer was composed of Tbl8Fe72Col0, having a Curie temperature of ca. 220~C, a compensation temperature below room temperature, and a coercive force of ca. 10 KOe.

3~a~

1 Example 24 A sample of magnetooptical disk was prepared in the same manner as in the Example 22, except that the second magnetic layer was composed of Tbl5Fe77Co8, having a Curie temperature of ca. 210~C, a compensation temperature below room temperature, and a coercive force of ca. 1 KOe.
The samples of the Examples 22, 23 and 24 were subjected to the test of recording and reprocution, according to the process of the Example 22 and with varying recording laser power. Tab. 11 summarizes the threshold values (where C/N ratio is saturated) of the recordings of first and second types, and the corre-sponding C/N ratios.
1;, Tab. 11 Sample1st type recording 2nd type recording , threshold v. C/N threshold v. C/N

Example 22 2 mW ~ 46 dB 6 mW 46 dB
Example 23 2.5 mW 49 dB 7 mW 49 dB

Example 24 4.5 mW 35 dB 6 mW 45 dB
__ Tab. 11 shows that the samples of the Examples 22 and 23 have higher sensitivities of the recording of first type and larger C/N ratios compared to the sample of the Example 24. These results reflect that the second magnetic layer has a compensation temperature higher than room temperature and increases the coercive force HL of the second magnetic layer il the course of loG l~Od.-~8 1 recording, whereby the second magnetic layer is stabilized with respect to the magnetic field and the inversion of magnetization of the first magnetic layer takes place stably from a low temperature.
s Also the comparison of the Examples 22 and 23 suggests that the Example 23 shows an increased C/N
ratio at the reproduction, though the recording sensitivity is somewhat lost, because of the higher Curie temperature of the first magnetic layer.

Protective layer Fig. 24 is a schematic cross-sectional view showing an embodiment of the magnetooptical recording medium of the present invention, wherein a protective 1~ layer 46 is provided on the second magnetic layer 3.
Said protective layer 46 is preferably formed with a thickness equal to or larger than 200 A, in order to form a continuous film. Also, as shown in Fig. 25, a protective layer 27 may be provided between the substrate 1 and the first magnetic layer 2, and said protective layer 47 is also preferably formed with a thickness equal to or larger than 200 A, in order to obtain a continuous film. The corrosion resistance of the medium can be further improved by adhering another substrate 49, as shown in Fig. 26, onto the protective layer 46 by means of an adhesive layer 48. Furthermore, recording and reproduction can be made from both sides if the 1 layers 47 to 46 are formed also on the substrate 49.
Said protective layer is composed of a dense non-magnetic material, usually an inorganic dielectric material, such as Si3N4, SiC, ZnS, AlN, SiO, A12O3, Si or Ge. In the structures shown in Figs. 25 and 26, the protective layers 46, 47 may be composed of different materials, but are preferably formed with a same material in consideration of the stability of the magnetic characteristics during storage. In Figs. 24 to 26, same components as those in Fig. 3 are represented by same numbers or symbols, and will not be explained further.

Example 25 1~ A pregrooved and preformatted glass disk substrate was set in a sputtering apparatus with ternary targets, and was rotated at a distance of 10 cm from the targets.
A GdTbFeAl alloy was sputtered, in argon gas, from a second target with a sputtering speed of 100 Almin. and a sputtering pressure of 5 x 10 Torr to obtain a first magnetic layer of Gdl5Tb5Fe79All of a thickenss of 210 A, TL = ca. 165~C and HH = ca. 10 KOe.
Then a TbFeCoCr alloy was sputtered, in argon gas, with a sputtering pressure 5 x 10 Torr to obtain a second magnetic layer of Tb23Fe60Col4Cr3 with a thick-ness of 500 A, TH = ca. 190~C and HL = ca. 1 KOe.

108 ~L,,;~ 4~

1 Subsequently a ZnS protective layer of 1500 A
in thickness was formed by sputtering, in argon gas, from a first target with a sputtering speed of 100 A/min. and a sputtering pressure of 5 x 10 Torr.
After these layer formations, the above-mentioned substrate was adhered to a polycarbonate plate with hot-melt adhesive material to obtain a magneto-optical disk. Said disk was mounted on a record/
reproducing apparatus, and was made to pass through, with a linear speed of ca. 5 m/sec., a unit generating a magnetic field of 2.5 KOe. Then the recording was conducted with a laser beam of a wavelength of 830 nm, concentrated to ca. 1 ~m and modulated in two levels of 4 and 8 mW, with a duty ratio of 50 % and a frequency ~f 500 KHz. The bias field was 100 Oe. Binary signals could then be reproduced by irradiation with a laser beam of 1.0 mW.
The above-explained experiment was repeated with a magnetooptical disk recorded already over the entire surface. The previously recorded signal components were not detected, and the possibility of over-writing was thus confirmed.

Example 26 Samples 26-1 to 26-11 of the structures shown in Tab. 11 were prepared in the same manner as in the Example 25. In Tab. 12, Mag(l) and Mag(2) indicate lo~ 3;~i~

1 the first and second magnetic layers prepared as explained in the Example 25 and composed respectively of Gd 5Tb5Fe79All and Tb23Fe60C~14 3 thicknesses and coercive forces explained before.
Also glass means a glass disk substrate. ~'ab. 12 shows the values of HL and ~W/2Msh measured on each sample.

Tab. 12 Sample Structure Protective HL(Oe) ~ /2MSh thickness 26-1 glass/Mag(l)/Mag(2) - 1000 800 26-2 glass/Mag(l~/Mag(2)/ z S 100 A 1000 400 Protect. layer 26-3 ibid. ZnS-200 1000 350 26-4 ibid. ZnS-400 1000 300 .) 26-5 ibid. ZnS-1000 1000 300 26-6 ibid. Si3N4 00 1000 300 26-7 ibid. SiC-200 1000 300 26-8 ibid. SiO-300 1000 300 26-9 ibid. S 3 4 1000 300 26-10 ibid. SiC-1000 1000 300 26-11 ibid. SiO-1000 1000 300 The values of HL and ~W/2Msh were measured by preparing a sample, composed of a first magnetic layer and a second magnetic layer on a substrate, in the same proce~s as in the Example 25, and determining the magnetic field inducing the inversion of the nagnetization I of the second magnetic layer, by means of the magneto-optical effect and in the presence of an external magnetic field.
In the following there will be explained the S relation, examined from the results shown in Tab. 12, between the protective layer and the condition HL ~
aw/2Msh for the stability of the record bit 4f shown in Fig. 5.
The sample 26-1 did not have a protective layer in contact with the second magnetic layer, and HL ~ aw/~Msh was 200 Oe. This indicates that the mutually opposite magnetizations of the first and second magnetic layers can stably exist with a margin of 200 Oe in the absence of the external field. In practice, however, 1 - 5 % of the magnetization of the second magnetic layer is inverted because inverted domains are generated by a smaller field.
The samples 26-2 to 26-5, having a ZnS
protective layer in contact with the second magnetic layer, showed a decrease in aw/2Msh with the increase in thickness of said protective layer. At a thickness of ZnS layer between 100 and 200 A, aw/2Msh reaches a substantially constant value same as that at a ZnS
thickness of 1000 A, providing a margin of ca. 700 Oe.
These samples did not show the above-mentioned inversion of magnetization of the second magnetic layer in the absence of the external field, and were therefore stable.

3; 3 1 Electron microscope observation of the samples proved that the protective layer was transformed from an island structure to a continuous film at a thickness of lO0 - 200 A. It is estimated that the protective layer generates a compression stress in the second magnetic layer, thus inducing a change in aw/2Msh.
In the samples 26-2 to 26-ll investigated was the effect of material constituting the protective layer formed in contact with the second magnetic layer.
It was confirmed that a protective layer composed of Si3N4, SiC or SiO had an effect of reducing aw/2Msh~ or of stabilizing the record bit.
When the samples 26-2 to 26-ll were maintained in an atmosphere of a temperature 45~C and a humidity of 65 % for 3 days, the sample 26-2 with the protective layer of lO0 A showed an increase of ~W/2Msh by about 30 %, other samples showed decreases less than lO %.

Example 27 A progrooved and preformatted polycarbonate disk substrate was set in a sputtering apparatus, with ternary targets, and was rotated at a distance of lO cm from the targets.
A Si protective layer of 500 A in thickness was formed by sputtering, in argon gas, from a first target with a sputtering speed of lO0 A/min and a sputtering pressure of 5 x lO Torr. Then a GdTbFeCo alloy was 1 1 ~ 1 3 ~

I sputtered, in argon gas, from a second target with a sputtering speed of 100 A/min. and a sputtering pressure of 5 x 10 3 Torr to obtain a first magnetic layer of Gdl0Tbl0Fe78Co2 with a thickness of 190 A, TL = ca.
5 160~C and HH = ca. 12 KOe.
Then a TbDyFeCo alloy was sputtered, in argon gas, with a sputtering pressure of 5 x 10 Torr to obtain a second magnetic layer of Tb12Dyl2Fe64Col2 of a thickness of 500 A, TH = ca. 185~C and HL = ca. l 0 KOe.
Then a Si protective layer of 1000 A in thick-ness was formed by sputtering, in argon gas, from the first target with a sputtering speed of 100 A/min. and a sputtering pressure of 5 x 10 Torr.

After these layer formations, the above-mentioned substrate was adhered to a polycarbonate plate with hot-melt adhesive material to botain a magnetooptical disk.
Said disk was mounted on a record/reproducing apparatus and was made to pass through, with a linear speed of ca. 8 m/sec., a unit generating a magnetic field of 2.5 KOe. The recording operation was then conducted with a laser beam of a wavelength of 830 nm, concentrated to ca. 1 ~m and modulated at two levels of 4 and 8 mW with a duty ratio of 50 % and a frequency of 2 MHz. The bias field was 100 Oe in a direction to facilitate the inversion of magnetization of the second magnetic layer.
Then binary signals could be reproduced by irradiation 1 3 ~ 0 ~3 rj ~3 1 with a laser beam of 1.5 mW.
The above-explained experiment was repeated with a magnetooptical disk already recorded over the entire surface. The previously recorded signal compo-5 nents were not detected in the reproduction, so that thepossibility of overwriting was confirmed.

Example 28 Samples 28-1 to 28-16 were prepared in the same manner as in the Example 27, as shown in Tab. 13. The first magnetic layer (Mag(l)) and the second magnetic layer (Mag(2)) were respectively composed of Gd Tb Fe78co2 and Tbl2DY12Fe64C~l2 nesses and coercive forces as in the Example 27. PC
indicates a polycarbonate substrate.
Tab. 13 shows the values of HL and ~W/2Msh measured on said samples.

~ . . ,, , . _ ~ " . , ~, I , . ..

- ~n O ~" O c~;, ~

Tab. 13 Sample Structure Prot. layer(l) Prot. layer(2) HL(Oe) aw/2Msh ~W/2Msh (Oe) material/ Omaterial/ O (oe) after storage thickness(A) thickness(A) 28-1 PC/Prot.Layer(l)/ ZnS/150 - 1200 1000 Mag(l)/Mag(2) 28-2 ibid. ZnS/1000 - 1000 800 28-3 PC/Prot.Layer(l)/ ZnS/800 ZnS/100 1000 400 Mag(1)/Mag(2)/
Prot.Layer(2) 28-4 ibid.ZnS/800 ZnS/200 1000 350 700 28-5 ibid.ZnS/800 ZnS/400 1000 300 300 28-6 ibid.ZnS/800 ZnS/800 1000 300 300 28-7 ibid.SiO/800 SiO/800 1000 300 300 28-8 ibid.SiO/800 SiO/800 1000 300 450 28-9 ibid. 3 4/ ~ 3 4/ 1000 300 300 28-10 ibid.SiC/800 SiC/800 1000 300 300 28-11 PC/Prot.Layer(l)/ ZnS/800 SiO/800 1000 300 400 Mag(1)/Mag(2)/ C
Prot.Layer(2) ~ o Tab. 13 (continued) Sample Structure Prot. layer(1) Prot. layer(2) HL(Oe) ~W/2Msh ~W/2Mlsh (Oe) material/ O material/ O (Oe) after storage thickness(A) thickness(A) 28-12 PC/Prot.Layer(1)/ ZnS/800 SiC/800 1000 300 450 Mag(l)/Mag(2)/
Prot.Layer(2) 28-13 ibid. ZnS/800 Si3N4/800 1000 300 500 28-14 ibid. Si3N4/800 ZnS/800 1000 300 500 . 28-15 ibid. Si3N4/800 SiC/800 1000 300 450 ~' - 28-16 ibid. Si3 4/ SiO/800 1000 300 450 iX~

1 3 ~j ~

1 The values of HL and ~W/2Msh were measured by preparing a sample, composed of a first magnetic layer and a second magnetic layer on a substrate, formed in the same manner as in the Example 27, and determining the magnetic field inducing the inversion of magnet-ization of the second magnetic layer, by means of the magnetooptical effect, in the presence of an external magnetic field.
The value of aw/2M5h after storage was obtained by placing the sa~ples 28-1 to 28-16 in an atmosphere of a temperature of 60~C and a relative humidity of 80 % for 500 hours. No results are shown for the samples 28-1 to 28-3, as the measurement was not possible by the deterioration of the magnetic characteristic of the second magnetic layer.
In the following there will be explained the relation, examined from the results shown in Tab. 13, between the protective layers provided on both sides of the first and second magnetic layers, and the condition HL ~ ~ /2MSh for the stability of the record bit 4f shown in Fig. 5.
The samples 28-1, 28-2 did not have a protective layer in contact with the second magnetic layer, and HL ~ ~W/2Msh = 200 Oe. This signifies that the mutually unstable arrangement of magnetizations of the first and second magnetic layers can stably exist with a margin of 200 Oe, in the absence of the external field. In 11 7 1 ~ ~ 0;

1 practice, however, 1-5 % of the magnetization of the second magnetic layer is inverted because inverted domains are generated by a smaller field.
The samples 28-3 to 28-6, having ZnS protective layers in contact with the first and second magnetic layers, showed a decrease in ~W/2Msh with the increase in the thickness of said protective layers. At a thickness of ZnS layers above ca. 200 A, aw/2Msh reaches a substantially constant value, providing a margin of ca. 700 Oe. These samples did not show the above-mentioned inversion of magnetization of the second magnetic layer in the absence of the external field, and the record bits were therefore stable.
Electron microscope observation of the samples proved that the protective layer was transformed from an island structure to a continuous film at a thickness of 100 - 200 A. It is estimated that the protective layer generates a compression stress in the second magnetic layer, thus inducing a change in ~W/2Msh.

In the samples 28-7 to 28-16, the protective layers were fixed to a thickness of 800 A at which the effect on ~W/2Msh is already saturated and which provides a sufficient protective effect for example for preventing the entry of moisture into the magnetic layers-In the sample 28-7 the sputtering pressure for SiO was 3 x 10 3 Torr for both first and second 118 ~ 30 i~
.

1 layers, while in the sample 28-8 the sputtering pressure for SiO was 8 x 10 3 Torr for the second layer only.
The sample 28-7 is a case in which the protective layers have a same stress, while the sample 28-8 is a S case in which the protective layers have mutually different stresses.
As a reference, the magnitude of stress was measured on samples for stress measurement, prepared by forming a SiO layer of ca. 5000 A by sputtering on a glass substrate of 1.5 mm in thickness. The layer formed with a sputtering pressure of 3 x 10 3 Torr provided a compression stress of ca. 70 kg/mm2, while that formed with a pressure of 8 x 10 3 Torr provided a compression stress of ca. 15 kg/mm2.
The samples 28-7 and 28-8 showed a value of aw/2Msh of ca. 300 Oe immediately after the preparation.
However, after a storage of 500 hours at 60~C and 80 %
relative humidity, the sample 28-8 with two SiO layers of different compression stresses showed an increase in aw/2Msh, rendering the record bits more unstable.
The samples 28-9 to 28-16 were prepared with various materials for the protective layers, but the sputtering pressures therefor were kept constant at 5 x 10 Torr.

All the samples showed a value of ~W/2Msh of ca. 300 Oe immediately after the preparation. Thus the recording is made stabler compared to the case without 1 the protective layers.
However, after a storage for 500 hours at 60~C
and 80 % relative humidity, the samples in which two protective layers are composed of different materials a showed a change in aw/2Msh.
The ZnS, SiC or Si3N4 sputtered layer of 5000 A in thickness, when subjected to the measurement of stress in the same manner BS the SiO layer, showed a compression stress of ca. 50, 80 or 90 kg/mm2 respectively. When stored for 500 hours at 60~C and 80 % relative humidity, these samples showed an increase in the stress, which was as high as ca. 20 to 30 ~ in SiC and Si3N4.
These results indicate that the stability of 1~ record bits is related to the stress of the protective layers. Particularly in consideration of prolonged storage, the balance of stresses of the protective layers should remain unchanged, and, for this purpose it is effective to prepare both protective layers with a same material, and with same manufacturing conditions such as the sputtering pressure.

Triple-layered structure The magnetooptical recording medium of the present invention may have a triple-layered structure, by adding a magnetic layer to the above-explained two-layered structure. Such triple-layered structure 13~O~

I will be explained further in the following.
Figs. 27 and 28 are schematic cross-sectional views of embodiments of the magnetooptical recording medium of the present invention. The medium shown in S Fig. 27 is composed of a pregrooved translucent substrate 51, and, a first magnetic layer 52, a second magnetic layer 53 and a third magnetic layer 54 lA~inAted thereon. The first magnetic layer 52 has a high Curie point (THl) and a low coercive force ~HLl), while the second magnetic layer has a low Curie point (TL2) and a high coercive force (HH2), and the third magnetic layer 54 has a high Curie point (TH3) and a low coercive force (HL3). The terms "high" and "low"
are defined by relative comparison between the first lS and third magnetic layers and the second magnetic layer, the comparison of coercive force being at room temper-ature.
Relation of magnitude of the Curie point or the coercive is not critically defined between the first and the third magnetic layers, but preferred relations are THl _ TH3 and HLl L3.
It is generally desirable that the first magnetic layer 52 has THl of 150 - 400~C and HLl of 0.1 - 1 KOe, the second magnetic layer 53 has TL2 of 70 - 200~C and HH2 of 2 - 10 KOe, and the third magnetic layer 54 has TH3 of 100 - 250~C and HL3 of 0 5 - 4 KOe.
In the magnetooptical recording medium of the I present invention, the neighboring magnetic layers are coupled by exchange force. The first and second magnetic layers are relatively strongly coupled, while the second and third are relatively weakly coupled.
In the magnetooptical recording medium of the present invention, the above-mentioned three magnetic layers are so coupled as to satisfy following relations:

w12 Ll (9) 2Msl 1 aw23 < H 3 (10) 2Ms3 3 L

wherein: aWl2 : magnetic wall energy of first and second 1~ magnetic layers 52, 53;
aW23 : magnetic wall energy of second and third magnetic layers 53, 54;
hl, h2, h3 : thicknesses of first, second and third magnetic layers 52, 53, 54;

MSl, MS2, Ms3 : saturation magnetizations of first, second and third magnetic layers.
The reason of these relations will be explained later.

The thickness, coercive force, saturation magnetization and magnetic wall energy of said magnetic layers 52, 53, 54 are so selected that the two states 1 of magnetization (shown by 60f in Fig. 29) of the finally recorded bits can exist in stable manner, or that the above-mentioned relations are satisfied.
Each magnetic layer can be composed of a substance exhibiting a vertical magnetic anisotropy and a magnetooptical effect, preferably an amorphous magnetic alloy of a rare earth element and a transition metal element such as GdCo, GdFe, TbFe, DyFe, GdTbFe, TbDyFe, GdFeCo, TbFeCo or GdTbCo.
Fig. 28 shows another embodiment of the magneto-optical disk of the present invention, in which protective layers 55, 56 are provided for improving the durability or enhancing the magnetooptical effect of the three magnetic layers 52, 53, 54.
An adhesive layer 57 is provided for adhering another covering substrate 58. Also recording and reproduction can be made from both sides if the covering substrate 58 with the layers from 55 to 56.
Now reference is made to Figs. 29 and 30 for explaining the recording process of the present embodi-ment, wherein, prior to the recording, the magnetization of the magnetic layer 52 and that of the magnetic layer 53 may be in a mutually parallel stable state, or in a mutually opposite (antiparallel) stable state.
The aforementioned apparatus shown in Fig. 6 can also be employed in the recording and reproduction of information to or from the magnetooptical recording ~ 3 '~ O ~ 8 1 medium of the present embodiment. It is assumed thata part of the magnetic layers has an initial magnet-ization as shown by 60a in Fig. 29. More specifically it is assumed, in this case, that the magnetizations 5 of the first, second and third magnetic layers are stable, prior to recording, when they are oriented in a same direction. The magnetooptical disk 9, being rotated by a spindle motor, passes the position of a magnetic field generating unit 8 shown in Fig. 6, 0 generating a magnetic field of which intensity is selected at a suitable level between the coercive forces of the second and third magnetic layers 53, 54 (magnetic field being upwards in the present embodiment), whereby, as shown by 60d in Fig. 29, the third magnetic layer 54 is uniformly magnetized while the second magnetic layer 53 retains the initial magnetization state. Also the first magnetic layer 52, strongly coupled with the second magnetic layer, retains the initial magnetization.

The rotated magnetooptical disk 9, in passing the position of a record/reproducing head 5, is irradiated by a laser beam with one of two power levels according to the signal from a recording signal generator 6. The first laser power is enough for heating the disk to a temperature close to the Curie point of the second magnetic layer 53, while the second laser power is enough for heating the disk to a 1 temperature close to the Curie point of the third magnetic layer 54. Referring to Fig. 30 showing the temperature dependence of the coercive forces of the magnetic layers 53, 54, the first laser power can heat 5 the disk close to TL2, while the second laser power can heat the disk close to TH3.
The first laser power heats the second and third magnetic layers 53, 54 close to the Curie point of the second magnetic layer 53, but the third magnetic layer 54 has a coercive force capable of stably maintaining the bit at this temperature. Thus, through a suitable selection of the recording bias magnetic field, a record bit shown by 60c in Fig. 29 can be obtained, as a preliminary recording of first 1~ type, from either state in 60b. The first magnetic layer 52 assumes the illustrated magnetization by the exchange coupling with the second magnetic layer 53.
The suitable selection of the bias magnetic field has the following meaning.

In said preliminary recording of first type, such bias magnetic field is essentially unnecessary, since the second magnetic layer 53 receives a force (exchange force) to orient the magnetization in a stable direction (same direction in this case) with respect to the direction of magnetization of the third magnetic layer 54. However said bias magnetic field is provided, in a preliminary recording of second 1 ~ 5 ~ 0 ~ ~

I type to be explained later, in a direction to assist the magnetic inversion of the third magnetic layer 54.
Also said bias magnetic field should preferably have a same magnitude and a same direction in the preliminary 5 recording with both first laser power and second laser power. In consideration of the foregoing, the bias magnetic field is preferably selected at a minimum necessary intensity required for the prel;m;~ry recording with the second laser power of the principle explained below.
When the disk is heated close to the Curie point of the third magnetic layer by a second laser power (preliminary recording of second type), the magnetization of said third magnetic layer 54 is inverted by the above-mentioned bias magnetic field.
Also the magnetizations of the second and first magnetic layers 53, 52 are oriented in a stable direction (same direction in this case) with respect to the magnetization of the third magnetic layer 54. In this manner a record bit as shown by 60b, in Fig. 29, can be from either state 60b.

Thus each area of the magnetooptical disk can have a record state 60c or 60d, shown in Fig. 29, by the bias field and by the first or second laser power corresponding to the input signal.
Then the magnetooptical disk 9 is further rotated, so that the record bit 60c or 60d passes again 0; ~

1 through the magnetic field generating unit 8, of which field intensity is selected, as explained before, between the coercive forces of the second and third magnetic layers 53, 54. .hus the record bit 60c remains unchanged S and assumes a state 60e. On the other hand, the record bit 60d causes the inversion of magnetization of the third magnetic layer 54 to assume a state 60f.
In order that the record bit 60f can exist stably, there are required the aforementioned relations:
2M lhl ~ HLl and w23 < H
s3 3 L3 for the following reasons.
lS a 12/2Mslhl indicates the magnitude of the exchange force received by the first magnetic layer, or represents the magnitude of a magnetic field acting to orient the magnetization of the first magnetic layer in a stable direction (same direction in the present case) with respect to the magnetization of the second magnetic layer. Therefore, in order that the magnetization of the first magnetic layer is always oriented in a stable direction (same direction in the present case) with respect to the magnetization of the second magnetic layer, the coercive force HLl of the first magnetic layer should be small~r than said exhange force, or ~wl2/2Mslhl > HLl.

.. . ..

1~7 1 Also, aw23/2Ms3h3 indicates the magnitude of the exchange force received by the third magnetic layer, or represents the magnitude of a magnetic field acting to orient the magnetization of the third 5 magnetic layer in a stable direction (same direction in the present case) with respect to the magnetization of the second magnetic layer. In order that the magnet-ization of the third magnetic layer is not inverted by said magnetic field, or in order that the record 60f, in Fig. 29, can stably exist, there should be satisfied ~ w23/2Ms3h3 ~ HL3 wherein HL3 is the coercive force of the third magnetic layer.
The second and third magnetic layers 53, 54 have to be exchange-coupled in order to obtain an effective bias magnetic field caused by the exchange force at the recording, but the above-mentioned condition cannot be satisfied, and the record bits 60f cannot exist stably, if the exchange coupling is too strong. In the preparation of the magnetooptical recording medium of the present invention, therefore, it is possible to optimize said exchange-coupling, by selecting the coercive force of the third magnetic layer 54 at a relatively large value within a range not exceeding the intensity of the magnetic field generated by the magnetic field generating unit, and, if said exchange coupling is excessively large, by suitably selecting the composition of the third magnetic layer 1 ~ 8 1 or by forming an intermediate layer of several to several thousands of Angstroms in thickness between the second and third magnetic layers (by exposing to gas or plasma reactive to the material of the second 5 magnetic layer after sputtering thereof or by sputtering a dielectric layer from a target).
The recording process of the present embodiment enables an overwriting operation, since the record bits 60e, 60f do not rely on the state prior to recording but solely on the laser power at the recording. The record bits 60e, 60f can be reproduced by irradiation with a reproducing laser beam and processing the resulting light with a signal reproducing unit 7 shown in Fig. 6. The magnitude ~modulation) of the reproduced signal principally depends on the magnetooptical effect of the first magnetic layer. This recording process can provide a recording with a reproduction signal of a high magnitude or a high degree of modulation because of the above-mentioned fact, and a fact that, in the medium of the present invention having three magnetic layers, a material with a high Curie point, or with a strong magnetooptical effect, can be used for the first magnetic layer 52 receiving the reproducing laser beam.

In the foregoing description relating to Fig.
29, it is assumed that the magnetizations of the first, second and third magnetic layers 52, 53, 54 are stable ~9 13'~ 0 1 when they are oriented in a same direction, but a similar process can be realized when the magnetizations are stable when they are oppositely oriented. Figs.
31 and 32 illustrate states of magnetization in such recording process, wherein 61a - 61f and 62a - 62f respectively correspond to the states 60a - 60f in Fig. 29, and the recording is conducted in a similar manner as shown in Fig. 29.

Example 29 A pregrooved and preformatted polycarbonate disk substrate was set in a sputtering apparatus with quaternary targets, and was rotated at a distance of 10 cm from the targets.

A ZnS protective layer of 800 A in thickness was formed by sputtering, in argon gas, from a first target with a sputtering speed of 100 A/min. and a sputtering pressure of 5 x 10 Torr.
Then a GdFeCo alloy was sputtered, in argon gas, from a second target with a sputtering speed of 100 A/min. and a sputtering pressure of 5 x 10 Torr to obtain a first magnetic layer of Gd20Fe56Co24 with a thickness of 400 A and THl = ca. 350~C. HLl of said first magnetic layer itself was lower than ca. 500 Oe, and the transition metal was prevailing in the sub-lattice magnetization.
Then a TbFe alloy was sputtered from a third 1 .~ 0 I target under similar conditions to obtain a second magnetic layer of Tbl8Fe82 with a thickness of 400 A
and TL2 = ca. 140~C. HH2 ~f said second magnetic layer itself was above ca. 5000 Oe, and the transition metal was prevailing in the sub-lattice magnetization.
Then a GdTbFeCo alloy was sputtered from a fourth target under similar conditions to obtain a third magnetic layer of Gdl3Tbl3Fe69 5Co4 5 with a thickness of 300 A and TH3 = ca- 210~C- HL3 ~f said third magnetic layer itself was ca. 500 - 1500 Oe, and the rare earth metal was prevailing in the sub-lattice magnetization.
Subsequently a ZnS protective layer of 2000 A
in thickness was formed by sputtering from the first target.
After these layer formations, the above mentioned substrate was adhered to a polycarbonate platè with hot-melt adhesive material to obtain a magnetooptical disk. Said disk was mounted on a record/
reproducing apparatus, and was made to pass through, with a linear speed of ca. 7 m/min., a unit generating a magnetic field of 2 KOe. Recording was then conducted with a laser beam of a wavelength of 830 nm, concentrated to ca. 1 ~m and modulated in two levels of 4 and 8 mW, with a duty ratio of 50 % and a frequency of 2 MHz.
The bias field was 150 Oe. Binary signals could be then reproduced by irradiation with a laser beam of 1 mW.

1 The above-explained experiment was repeated with a magnetooptical disk already recorded over the entire surface. The previously recorded signals were not detected in the reproduction, and the possibility 5 of overwriting was thus confirmed.
The exchange force ~w23/2Ms3h3 ~f the second and third magnetic layers can also be regulated for example by varying the time between the formation of the second magnetic layer and the formation of the third magnetic layer.
By varying said time between the formations of the second and third magnetic layers from 30 sec.
to 30 min. and further to 3 hours, the value of ~W23/2Ms3h3 was reduced from l.0 KOe to 0.7 KOe and further to 0.4 KOe.

Example 30 Samples of magnetooptical disk were prepared, as shown in Tab. 14, maintaining the thicknesses and the coercive forces of the Example 29 for the protective layers and the magnetic layers, but varying the ratio of the transition metal and the rare earth element in each of three layers.
In each sample, the stability of the record bits 60e, 60f was investigated through the measurement of the applied external magnetic field required for inverting the magnetization of the magnetic layers.

1 3~ ~ J~g~

1 Also recording test was made in the same method as in the Example 29, and the state of record of 4 and 8 mW was evaluated, as shown in Tab. 14.

,, 25~

". ,~, , . , ",, , ~ , .. . ..

~j Tab. 14 Composition Recording bit Record state Sample 1st mag. layer 2nd mag. layer 3rd mag. layer stability rich in rich in rich in 60e 60f 4mW 8mW
Ex. 29 trans. metal trans. metal rare earth + + + +
30-1 rare earth rare earth trans. metal + + + +
30-2 trans. metal trans. metal trans. metal + - - +
30-3 trans. metal rare earth rare earth + - - +
30-4 rare earth trans. metal trans. metal + - - + C~
30-5 rare earth rare earth rare earth I + - - +
3~-6 rare earth trans. metal rare earth + + + +
30-7 trans. metal rare earth trans. metal + + + +

o 1;~4 lX ~ 0~

1 The stability of record bits is represented by "+" or "-" respectively if the state 60e or 60f can stably exist or not in the absence of the external magnetic field.
The record state was evaluated as "-" if the reproduction signal could not be confirmed from the record with 4mW or 8mW, "+" if the reproduction signal could be confirmed but not satisfactory, and "+"
if a satisfactory reproduction signal of C/N ratio of about 40 dB or higher could be obtained.
The results in Tab. 14 indicates that, for achieving satisfactory recording with stable record bits, the first and second magnetic layers should prefer-ably be rich simultaneously in the transition metal (transition metal prevailing in the sub-lattice magnet-ization) or in the rare earth element (rare earth element prevailing in the sub-lattice magnetization) compared to the compensation composition, and the third magnetic layer should preferably prevailing element opposite to that in the first and second magnetic layers, as in the samples 29 and 30-1.
In another embodiment of the triple-layered magnetooptical recording medium of the structure shown in Figs. 27 and 28, there may be adopted relations:

H1 ' H3 > H2 (11) T ' T2 ' T (12) wherein Tl, T2 and T3 are the Curie points respectively 13~ ~3-l~O.i~

I of the first, second and third magnetic layers 52, 53, 54, and Hl, H2 and H3 are the coercive forces re-spectively of said magnetic layers.
In general, the first magnetic layer 52 has 5 Tl in a range from 70 to 200~C and Hl in a range from 2 to 10 KOe, while the second magnetic layer 53 has T2 from 90 to 400~C and H2 from 0.1 to 1 KOe, and the third magnetic layer 54 has T3 from 150 to 400~C and H3 from 0.5 to 4 KOe. In such case, in order to achieve stable recording, the layer thicknesses, saturation magnet-izations and magnetic wall energies should preferably be so regulated as to substantially satisfy following relations:

awl2 sl 1 1 (13) Ms3 3 (14) ~w12 w23 (15) 2MS2h2 2MS2h2 wherein awl2 is the magnetic wall energy of the first and second magnetic layers; ~w23 is the magnetic wall energy of the second and third magnetic layers; hl, h2 and h3 are the thicknesses of the first, second and third magnetic layers; and MSl, MS2 and Ms3 are the saturation magnetizations of said layers.

The information recording is conducted in the following manner, with the apparatus shown in Fig. 6.

1 3 ~ 3 ~; ~

1 It is assumed, prior to the recording, that the magnetizations of the first magnetic layer 52 and of the second magnetic layer 53 are in a mutually opposite (antiparallel) stable state, while the magnetizations of the second and third magnetic layers 53, 54 are in a mutually parallel stable state. It is also assumed that a part of the magnetic layers is in a state of magnetization shown by 63a in Fig. 33.
The magnetooptical disk 9, being rotated by the spindle motor as shown in Fig. 6, passes through the position of the magnetic field generating unit 8, generating a magnetic field, which is upwards in the present case and of which intensity is adjusted to a value between the coercive forces of the first and second magnetic layers 52, 53, whereby the second and third magnetic layers 53, 54 are uniformly magnetized but the first magnetic layer 52 retains the initial state of magnetization, as shown by 63b in Fig. 33.
Then the magnetooptical disk 9 is rotated, and, in passing the record/reproducing head 5, is irradiated by a laser beam with one of two power levels, according to a signal from the recording signal generator 6. The first laser power is enough for heating the disk to a temperature close to the Curie point of the first magnetic layer (a temperature close to Tl and allowing to orient the magnetization of the first magnetic layer in a stable direction with respect to the magnetizations 137 1 3 A~d' 1 of the second and third magnetic layers), while the second laser power is enough for heating the disk to a temperature close to the Curie points of the second and third magnetic layers (a temperature close to T2 and T3 a and allowing to uniformly invert the magnetizations of the second and third magnetic layers). Referring to Fig. 34 showing the temperature dependence of the co-ercive forces of the magnetic layers 52, 54, the first laser power can elevate the te~perature of the disk close to Tl while the second laser power can elevate the temperature of the disk close to T3.
The first laser power heats the first and third magnetic layers 52, 54 close to the Curie point of the first magnetic layer 52, but the third magnetic layer 54 has a coercive force allowing to stably maintain the record bit at this temperature. Thus, through a suitable selection of the bias magnetic field, a record bit as shown by 63c in Fig. 33 can be formed, as a preliminary recording of first type, from either state of magnetization shown by 63b. The magnetization of the ~second magnetic layer, having a stronger exchange force with the third magnetic layer than with the first magnetic layer, becomes always parallel with the magnetization of the third magnetic layer, as illustrated.

The suitable selection of the bias magnetic field has the following meaning.

In said preliminary recording of first type, the ~ .. . , ~ . . .. . . .

1 3 8 ~ Q .~

1 bias magnetic field is essentially unnecessary because the first magnetic layer 52 receives an exchange force to orient the magnetization thereof in a stable (opposite or antiparallel in the present case) direction S with respect to the magnetization of the third magnetic layer 54. However the bias magnetic field is provided for the preliminary recording with the second laser power to be explained later, in a direction to assist the inversion of magnetization of the third magnetic layer 54, or a direction to hinder the preliminary recording of first type. It is convenient that said bias magnetic field is maintained in a same intensity and in a same direction, both in the preliminary recording of fir~t and second types. In consideration of these conditions, the bias magnetic field should preferably be maintained at a minimum necessary level for the preliminary recording of second type to be explained in the following, and such bias field corresponds to the above-mentioned suitable selection.
When the disk is heated close to the Curie point of the third magnetic layer 54 by the second laser power for the preliminary recording of second type, the magnetizations of the third and second magnetic layers 54, 53 are inverted by the bias magnetic field selected as explained above. Subsequently the magnetization of the first magnetic layer 52 is also oriented in a stable (opposite or antiparallel in the present case) direction 130 1~4~

1 with respect to the magnetization of the third magnetic layer 54. In thi~ ~nner a preliminary recording bit as shown by 63d in Fig. 33 can be formed from either state of magnetization shown by 63b.
In this manner each area of the magnetooptical disk can form a preli~i n~ry record of the state 63c of 63d as shown in Fig. 33, by means of the bias magnetic field and the laser power of first or second level according to the input signal.
Then the magnetooptical disk 9 is further rotated and passes again the position of the magnetic field generating unit 8, generating a magnetic field of which intensity is selected between the coercive forces of the first and third magnetic layers 52, 54 as 3 explained before, whereby the record bit 63c remains unchanged as a final record bit 63e, while the record bit 63d assumes another final record state 63f as the result of magnetic in~ersion of the third magnetic layer.

In order that the record bit 63f can stably exist, the aforementioned conditions (13) to (15) should be satisfied, because of the following reasons.
~Wl2/2Mslhl indicates the magnitude of the exchange force received by the first magnetic layer, or represents the magnitude of a magnetic field acting to rearrange the magnetization of the first magnetic layer in a stable (opposite or antiparallel in the 140 ~ 3 ~ 00.;,~
....

1 present case) direction with respect to the magnet-ization of the second magnetic layer. Therefore, in order that the first magnetic layer can retain its magnetization unchanged against said magnetic field, said layer should have a coercive force Hl larger than the magnitude of ~aid exchange force (awl2/2Mslhl <

Hl .
The second magnetic layer 53 receives, from a magnetic wall at the interface with the first magnetic layer 52, an exchange force of a magnitude awl2/2Ms2h2 and of a direction to orient the magnetization of the second magnetic layer in a stable direction with respect to the magnetization of the first magnetic layer 52, and also receives, from a magnetic wall at the interface with the third magnetic layer 54, an exchange force of a magnitude aw23/2Ms2h2 and of a direction to orient the magnetization of the second magnetic layer in a stable direction with respect to the magnetization of the third magnetic layer 54.
Consequently, in order that the magnetization of the second magnetic layer is constantly oriented in a stable direction with respect to the magnetization of the third magnetic layer during and after the recording operation, there are required conditions (aw23/2Ms2h2) ~ (a 12/2MSlhl) and (aW23/2Ms2 2 2 Also aw23/2Ms3h3 indicates the magnitude of the exchange force received by the third magnetic layer 14~ ~ 100;i3 1 54, or represents the magnitude of a magnetic field acting, across the second magnetic layer 53, to rearrange the magnetization of the third magnetic layer in a stable (opposite or antiparallel in the present case) direction with respect to the magnet-ization of the first magnetic layer 52. In order that the magnetization of the third magnetic layer 54 is not inverted again~t said magnetic field, or in order that the record bit 63f in Fig. 33 can stably exist, thecoercive force H3 of the third magnetic layer 54 should satisfy a condition (aw23/2Ms3h3) < H3.
From the foregoing explanation the interfacial magnetic wall energies should satisfy a condition aW23 ~ awl2' so that a) it is desirable to realize a weak coupling between the first and second magnetic layers, by adopting a composition rich in rare earth element compared with the compensation composition for one of said layers and a composition rich in transition metal for the other of said layers; and b) to realize a strong coupling between the second and third magnetic layers by adopting compositions both rich in the rare earth element or compositions both rich in the transition metal.

Also it is possible to reduce the exchange force from the third magnetic layer to the second magnetic layer, by reducing the exchange force awl2/2Ms2h2 from ,.. . , ~ . . . .. .. . .. . . ..

1 4 2 i 3 ~ 3 ~

1 the first magnetic layer to the second magnetic layer, and the exchange force ~W23/2M~2h2 from the third magnetic layer to the second magnetic layer. For this purpose it is advantageous to reduce the saturation S magnetization MS2 Qf the second magnetic layer, namely to mi n;m; ze the coercive force H2 thereof. Therefore at least a condition H3'> H2 is desirable.
In the recording process of the present embodi-ment, the record bits 63e and 53f do not depend on the state prior to recording but are solely controlled by the recording laser power, so that an overwriting operation is possible. The record bits 63e, 63f can be reproduced by irradiation with a reproducing laser beam and by processing of the resulting light with the recording signal reproducer 7.

Example 3l ~ A pregrooved and preformatted polycarbonate disk substrate was set in a sputtering apparatus with four targets, and was rotated at a distance of lO cm from the targets.
A ZnS protective layer of 800 A in thickness was formed by sputtering, in argon gas, from a first target, with a sputtering speed of lO0 A/min. and a sputtering pressure of 5 x lO Torr.
Then a TbFe alloy was sputtered, in argon gas, from a second target with a sputtering speed of 14 3 ~ r~ 1 Q ~) ~1 t 3 1 1~0 A/min. and a sputtering pressure of 5 x 10 Torr to obtain a first magnetic layer of Tbl8Fe82 of a thickness of 300 A and Tl = ca. 140~C. Hl of said first magnetic layer itself was ca. 10 KOe, and the transistion metal was prevailing in the sub-lattice magnetization.
Then a TbFeCo alloy was sputtered under similar conditions from a third target to obtain a second magnetic layer of Tb25Fe65Col0 with a thickness of 150 A and T2 = ca. 210~C. H2 ~f said second magnetic layer itself was lower than ca. 200 Oe, and the rare earth element was prevailing in the sub-lattice magnetization.
Subsequently a TbFeCo alloy was sputtered under similar conditions from a fourthtarget to obtain a third magnetic layer of Tb23Fe67Col0 with a thickness of 300 A and T3 = ca. 210~C. HL3 ~f said third magnetic layer itself was ca. 500 - 1500 Oe, and the rare earth element was prevailing in the sub-lattice magne~tization.
Then a ZnS protective layer of 2000 A in thickness was formed by sputtering from the first target under s~ Ar conditions.
After said layer formation, the above-mentioned substrate was adhered to a polycarbonate plate with hot-melt adhesive material to obtain a magnetooptical disk. Said disk was mounted on a record/reproducing apparatus, and was ~ade to pass through, with a linear speed of ca. 7 m/sec., a unit for generating a magnetic 144 ~ 0 1 field of 2 KOe. The recording was conducted with a laser beam of a wavelength of 830 nm, concentrated to ca. 1 ~ and modulated in two levels of 4 and 8 mW
with a duty ratio of 50 % and a frequency of 2 MHz.
5 The bias field was 150 Oe. Then binary signals could be reproduced by irradiation with a laser beam of 1 mW.
The above-explained experiment was repeated with a magnetooptical disk already recorded over the entire surface. The previously recorded signal component~ were not detected in the reproduction, and the possibility of overwriting was thus confirmed.

Example 32 Samples of magnetooptical disk shown in Tab. 15 were prepared in the same manner as in Example 31.
The thickness of the layers were maintained same as in the Example 31 for the purpose of comparison, but the compositions of the first, second and third magnetic layers were selected in various combinations of compositions rich in the rare earth element and those rich in the transition metal. The coercive forces of the first and third magnetic layers were respectively regulated to ca. 10 KOe and ca. 1 KOe through the control of contents of Tb and Fe.
In Tab. 15, TM means a composition rich in the transition metal compared with the compensation 1 34'~G~

1 composition, and RE means a composition rich in the rare earth element.

Tab. 15 Sample Composition Bit Sta- Record 1st mag. 2nd mag. 3rd mag. H2 of 2nd bility state layer layer layer mag. layer rich in rich in rich in EX.31 TM RE RE 200 Oe + +
3Z-1 TM RE RE 500 + +

32-2 TM RE RE 1000 + +

32-4 RE TM TM 300 + +
32-5 RE TM TM 700 + +
32-6 TM TM RE 500 +

32-8 RE RE TM 200 +

2~

146 L 3 ~ o a ~ ~

I In the first magnetic layer, the TM-rich composition wàs Tbl8Fe82, while the RE-rich composition was Tb20FE80. In the third magnetic layer, the TM-rich composition was Tbl5FE75Co10, while the RE-rich composi-tion was Tb23Fe67C~10 The stability of the record bit, particularly in the state 63f, was investigated on thus prepared samples, by measuring the external magnetic-field inducing the inversion of magnetization by a VSM. In 10 the results shown in Tab. 15, ~+ n indicates stable existence of the record bit in the state of 63f in the absence of external field, while ~-~ indicates a partial inversion of the magnetization of the third magnetic layer. Also the recordings of first and second types 15 were conducted in the same method as in the Example 31, and the result was evaluated as ~+u or ~-~ respectively if satisfactory reproduction signal was obtained or not.
. The results on the Example 31 and the Example 32-1 to 32-5 listed in Tab. 15 indicate that satis-20 factory bit stability and satisfactory record statecould be obtained, as explained in the foregoing, by a combination of a TM-rich first magnetic layer with RE-rich second and third magnetic layers or of an RE-rich first magnetic layer with TM-rich second and third 25 magnetic layers, and when the coercive force H2 of the second magnetic layer is smaller than the coercive force H3 of the third magnetic layer.

l~llOd-1 Example 33 Samples of magnetooptical disk shown in Tab. 16 were prepared in the same manner as in the Example 31.
The thicknesses of the layers were maintained same as in the Example 31 for the purpose of comparison, but the composition of the second magnetic layer was varied in comparison with the Example 31. The coercive force thereof was less than ca. 200 Oe, and the rare earth element was prevailing in the sub-lattice magnetization, 1() as in the Example 31.
The prepared samples were evaluated in the same manner as in the Example 32.

Tab. 16 Composition of Bit Record Sample 2nd mag. layer T2(~C) stability state 33-1 24 76 -140 +
33-2 13 12 75 160 + +
33_3 Tbl3Gd12Fe73C 2180 + +
33-4 Tbl3Gdl2Fe71Co4200 + +
33-5 DY2sFe75 70 + +
33-6 Tb5 Gd20Fe7o d 250 +

33-7 Tb5 Gd20Fe65C~10 The results of the samples listed in Tab. 16 indicate that a satisfactory recording could be obtained 1 4 8 ~ 3 ~
.,~

1 when the curie point T2 of the second magnetic layer is located between Tl and T3, (Examples 33-1 to 33-4), the recording of first type could not be satisfactorily made if T2 was lower than Tl (Example 33-5).
Also the recording of second type could not be made satisfactorily when T2 is higher than T3 (Examples 33-6, 33-7).
Quadraple-layered structure Furthermore, the magnetooptical recording medium 10 of the pre8ent invention may be provided with four magnetic layers, as will be explained in the following.
Figs. 35 and 36 are 6chematic cross-sectional views of quadraple-layered embodiments of the magneto-optical recording medium of the present invention. The 15 medium shown in Fig. 35 is composed of a pregrooved translucent sub~trate 71, and, a first magnetic layer 72, a second magnetic layer 73, a third magnetic layer 74 and a fourth magnetic layer 75 l~m; nAted thereon.
Said first, second, third and fourth magnetic layers 20 respectively have Curie points Tl, T2, T3 and T4;
coercive forces Hl, H2, H3 and H4; thicknesses hl, h2, h3 and h4; and saturation magnetizations MSl, MS2, M 3 and M 4. Also the magnetic wall energies between the first and second magnetic layers, between the second and 25 third magnetic layers, and between the third and fourth magnetic layers are respectively represented by ~w12 ~w23 and ~w34 The above-mentioned four magnetic 149 ~ 3~0~

1 layers satisfy following conditions I - IV, through mutual exchange-couplings.
I. For the Curie points of the magnetic layers:
Tl, T4 > T2, T3 (16) II. For the coercive forces of the magnetic layers:
H2 ~ H4 > Hl, H3 (17) III. For the thicknesses of the magnetic layers:
hl + h2 ~ 250 A

hl + h2 + h3 + h4 > 600 ~ (18) IV. For the saturation magnetizations, thick-nesses, coercive forces and magnetic wall energies of the magnetic layers:

w12 > H (19) lS sl 2Ms2 2 52 2 (20) 32MS3h3 ' 2M 3h3 (21) H ~w34 42M 4h4 (22) It is generally desirable that the first magnetic layer 72 has Tl of 150 - 400~C, Hl of 0.1 - 1 KOe, a thickness hl of 50 - 300 A; the second magnetic layer 73 has T2 of 70 - 200~C, H2 of 2 - 12 KOe, a thickness h2 of 50 - 300 A; the third magnetic layer 150 13~

1 74 has T3 of 0 - 200~C, H3 of 0.l - l KOe, a thickness h3 of 50 - 300 A and the fourth magnetic layer 7S
has T4 of l00 - 300~C, H4 of 0.5 - 4 KOe, and a thickness h4 of 50 - 600 A.
Each magnetic layer can be principally composed of a substance exhibiting a vertical magnetic anisotropy and a magnetooptical effect, preferably an amorphous alloy of a rare earth element and a transition metal element, such GdCo, GdFe, TbFe, DyFe, GdTbFe, TbDyFe, 10 GdTbFeCo, TbFeCo or GdTbCo.
In the magnetooptical recording medium of the present invention, neighboring magnetic layers are coupled with an exchAnge force. The first and second magnetic layers are relatively strongly couple, while 15 the second and third magnetic layers relatively weakly coupled, and the third and fourth magnetic layers are relatively woakly coupled. A large magnetic wall energy is present between the strongly coupled layers, while a weak magnetic wall energy is present betwsen 20 the weakly coupled layers. The value of the magnetic wall energy ~w is optimized in regulating Ms, h, H and ~w for satisfying the aforementioned relations IV.
In the four magnetic layers 72, 73, 74, 75 the thickness, coercive force, saturation magnetization 25 and magnetic wall energy of each layer are so regulated to satisfy the aforementioned relations, and such regulation will stabilize the finally completed two 15~ 1340d ,~
~.

1 states ,of magnetization shown by 70e and 70f in Fig. 37.
In another embodiment of the magnetooptical recording medium shown in Fig. 36, protective layers 76, 77 are provided for improving the durability of 5 the four magnetic layers 72, 73, 74, 75 or enhancing the magnetooptical effect thereof.
An adhesive layer 78 i8 provided for adhering a covering substrate 79. Recording and reproduction can be made from both sides if the layers of 72 to 77 are formed also on said covering substrate 79.
Now reference is made to Figs. 37 and 38 for explaining the recording process of the present embodiment, wherein, prior to the recording, among four magnetic layers, the magnetizations of mutually neighboring layers may be in a mutually parallel stable state, or in a mutually opposite (antiparallel) stable state.
However, since mutually parallel magnetizations provide a relatively strong coupling while mutually antiparallel magnetizations provide a relatively weak coupling, the magnetooptical recording medium of the present invention should preferably have such state of magnetization that the magnetizations of the first and second magnetic layers are in a mutually parallel stable state while those of the second and fourth magnetic layers are in a mutually antiparallel stable state.
The recording and reproducing operation on the 152 13 1Q~

1 medium of the present embodiment can also be conducted with the apparatus shown in Fig. 6. It is assumed that a part of said medium has an initial magnetization as shown by 70a in Fig. 37, and that the magnetizations S of the third and fourth magnetic layers are in a mutually parallel stable state. As shown in Fig. 6, the magnetooptical disk 9, being rotated by a spindle motor, pas~es the position of a magnetic field generating unit 8, generating an upward magnetic field of which 10 intensity is selected at a suitable level between the coercive forces of the second and fourth magnetic layers 73, 75, whereby, as shown by 70b in Fig. 37, the fourth magnetic layer 75 is uniformly magnetized, and the third magnetic layer 74, strongly coupled with the fourth layer 75, is also magnetized in the same direc-tion. On the other hand, the second magnetic layer 73 retains the initial magnetization, and the first magnetic layer 72, ~trongly coupled with the second magnetic layer 73, also retains the initial magnetization.

The rotated magnetooptical disk 9, in passing the position of a recordJreproducing head 5, is irradiated by a laser beam with one of two power levels according to the signal from a recording signal generator 6. The first laser power is enough for heating the disk to a temperature close to the Curie point of the second magnetic layer 73, while the second laser power is enough for heating the disk to a 15~
....

1 temperature close to the Curie point of the fourth magnetic layer 75. Referring to Fig. 38 showing the temperature dependence of the magnetic layers 73, 75, the first laser power can heat the disk close to T2, 5 while the second laser power can heat the disk close to T4.
The first laser power heats the second and fourth magnetic layers 73, 75 clo6e to the Curi¢ point of the second magnetic layer, but the fourth magnetic 10 layer 75 has a coercive force capable of stably maintaining the bit at this temperature. Thus, through a suitable selection of the recording bias magnetic field, the magnetizations of the magnetic layers are oriented in stable directions with respect to the 15 magnetization of the fourth magnetic layer in the course of temrerature lowering of the record bit area.
In this m~nner a record bit shown by 70c in Fig. 37 can be formed, as a prel;~;nAry recording of first type, from either state of magnetization shown 20 by 7Ob-The suitable selection of the bias magnetic field has the following meaning. In said preliminary recording of first type, such bias magnetic field is essentially unnecessary, since the magnetizations of the first, second and third magnetic layers receive forces (exchange forces) to orient the magnetizations thereof in respective stable directions with respect 1 5 4 ~ O O ~ ~

1 to the magnetization of the fourth magnetic layer 75.
However said bias magnetic field is provided, in a prel;~;n~ry recording utilizing the second laser power to be explained later, in a direction to assist the 5 magnetic inversion of the fourth magnetic layer 75.
Also said bias magn-tic field should preferably have a ~ame magnitude and a same direction in both prelimi-nary recordings with the first and ~econd laser power.
In consideration of such conditions, the bias magnetic 10 field is preferably selected at a minimum nece6sary intensity reguired for the prelimin~ry recording with the second laser power to be explained later, and such selection corresponds to the above-mentioned suitable selection.

When the disk is heated close to the Curie point of the fourth magnetic layer 75 by a second laser power (prel;~;n~ry recording of second type), the magnetization of said fourth magnetic layer 75 is inverted by the above-mentioned bias magnetic field.
20 Al~o the magnetizations of the first, second and third magnetic layers are oriented in stable directions with respect to the magnetization of the fourth layer 75.
Thu6 a record bit shown by 70d in Fig. 37 can be formed from either state of magnetization shown in 70b.

Thus each area of the magnetooptical disk can have a preliminary record bit 70c or 70d, shown in Fig. 37, by the bias field and by the first or 155 134~0~

I second laser power corresponding to the input signal.
Then the magnetooptical disk 9 is further rotated, 80 that the prelimin~ry record bit 70c or 70d passe6 again through the magnetic field generating unit 8, of which field intensity is selected, as explained before, between the coercive forces of the magnetic layer6 73, 75. Thus the record bit 70c remains l~n~h~nged and assumes a final record state 70e.
On the other hand, the record bit 70d causes the 10 inversion of magnetizations of the third and fourth magnetic layers 74, 75 to assume another final record state 70f.
The aforementioned relations (18) to (22) are required in order that the record bit 70f can stably 15 exist.
The relations (20) and (22) are required in order that the second and fourth magnetic layers assume an un~table state.
The relation (19) iB required to cause a 20 strong coupling between the first and second magnetic layers, and to constantly orient the magnetization of the first magnetic layer in a stable direction with respect to the magnetization of the second magnetic layer.
The relation (21) is required in order to constantly stabilize the magnetization of the third layer with respect to the magnetization of either the 15g 1~0.~'~

1 fourth or the second magnetic layer.
Also the relation (18) is required in order to optimize the recording sensitivity and the C/N
ratio in reproduction, as will be clarified in the 5 following examples.
The record bit states 70e, 70f do not depend on the states prior to recording but are solely determined by the recording laser power, so that the over-writing operation is rendered possible. The record bits 70e, 10 70f can be reproduced by irradiating the disk with a reproducing la~er beam and processing the resulting light with the recording signal reproducer 7.
The intensity or modulation of the reproduced signal principally depends on the magnetooptical effect 15 of the first and second magnetic layers a~ explained before. The present embodiment enables recording with strong reproduction signal, since a material of a high Curielpoint, namely of a large magnetooptical effect, can be used for the first magnetic layer which receives 20 the reproducing beam.
Example 34 A pregrooved and preformatted polycarbonate disk substrate was set in a sputtering apparatus, with ternary targets, and was rotated at a distance of 25 10 cm from the targets.
A ZnS protective layer of 900 A in thickness was formed by sputtering, in argon gas, from a first 1 target with a sputtering speed of 100 A/min. and a sputtering pressure of 5 x 10 3 Torr.
Then a GdFECoTi alloy was sputtered, in argon gas, from a second target with a sputtering S speed of 100 A/min. and a sputtering pres6ure of 5 x 10 Torr to obtain a first magnetic layer of Gdl8Fe55Co24Ti3 with a thickness of 200 ~ and Tl =
ca. 350~C. Hl of said first magnetic layer itself was lower than ca. 300 Oe, and the transition metal 10 was prevailing in the sub-lattice magnetization.
Then a TbFe alloy was sputtered from a third target under simi lAr conditions to obtain a second magnetic layer of Tbl8Fe82 with a thick-ness of 150 A and T2 = ca. 140~C. H2 of said 15 second magnetic layer itself wa~ 10 KOe, and the transition metal was prevailing in the sub-lattice magnetization.
~ Then a GdTbFe alloy was sputtered from a fourth target under similar conditions to obtain 20 a third magnetic layer of GdllTbllFe76 with a thick-ness of 100 A and T3 = ca. 160~C. H3 of said third magnetic layer was 100 - 300 Oe, and the rare earth element was prevailing in the sub-lattice magnetization.
Then a TbFeCo alloy was sputtered from a fifth target under similar conditions to form a fourth magnetic layer of Tb23Fe72Co5 with a thickness of 200 15~ o ~s ,~

1 and T4 = ca. 210~C. H4 of 6aid fourth magnetic layer itself was 500 - 1500 Oe, and the rare earth element was prevailing in the sub-lattice magnetization.
Subsequently a ZnS protective layer of 2000 A
in thickne6s was formed by ~puttering from the first 2 target under the same conditions as before.
After said layer formations, the above-mentioned substrate was adhered to a polycarbonate plate with hot-melt adhe6ive material to obtain a magnetooptical 10 disk.
Said disk was mounted on a record/reproducing apparatus, and was made to pa6s through a unit generating a magnetic field of 2 KOe, with a linear speed of ca. 7 m/6ec. Recording was condu¢ted with a laser beam of a 15 wavelength of 830 nm, focused to ca. 1 ~m and modulated in two levels of 4 and 8 mW, with a duty ratio of 50%
and a frequency of 2 MHz. The bias field was 150 Oe.
Binary signals could then be reproduced by irradiation with a laser beam of 1 mW.

The above-explained experiment was repeated with a magnetooptical disk already recorded over the entire surface. The previously recorded signals were not detected in the reproduction, and the possibility of overwriting was confirmed in this manner.

25~ EXample 35 \ Samples of magnetooptical disk was prepared ~ith the same materialG and Gtructure aG in the Example 1 34, with varying thicknesses in the first to fourth magnetic layers. Results of measurements of reflec-tance and Kerr rotation angle are shown in Tab. 17.
The reflectance was generally in a range of 23 - 24%, including the reflection of ca. 4~ from the surface of the substrate. Also a larger Kerr rotation angle will provide a larger C/N ratio of the reproduction signal.
Tab. 17 hl h2 h3 h4 hl+h2 hl+h2+h3+h4 Ok Sample (A) (A) (A) (A) (A) (A) (~~
35- 150 100 100 700 150 950 0.70 35- 2100100 100 700 2001000 0.76 35- 3150100 100 700 2501050 0.83 35- 4200100 100 700 3001100 0.83 35- 5100 50 100 700 150 950 0.72 35- 6100100 100 700 200100~ 0.76 35- 7]00150 100 700 2501050 0.79 35- ~100200 100 700 3001100 0.79 35- 9100300 100 700 4001200 0.79 35-10200100 50 100 300 450 0.60 35-11200100 100 100 300 500 0.74 35-12200100 100 200 300 600 0.83 35-13200100 100 300 300 700 0.83 35-14200100 100 400 300 800 0.83 35-1520010~ ]oo 500 300 900 0.83 35-'6200100 100 600 3001000 0.83 1~0 1~ ~Q i).-i~

1 The results on the samples 35-1 to 35-9 indicate that the Kerr rotation angle Ok shows ~aturation close to a value for a sufficiently thick first magnetic layer, if the total thickness hl + h2 of the first and second magnetic layers ~xceeds 250 A, only when the total thickness of the first to fourth magnetic layer is sufficiently large.
Also the results on the samples 35-10 to 35-16 indicate that the Kerr rotation angle Ok shows satura-10 tion, if the total thickness hl + h2 of the fir6t andsecond magnetic layers exceeds 250 A, when the total thickness of the first to fourth magnetic layers is equal to or larger than 600 A.
Therefore, in order, to obtain a high C/N ratio lS at the reproduction, it is preferable to select the layer thicknesse~ in such a ~nner as to satisfy s hl + h2 ~ 250 A and hl + h2 + h + h 600 ~
~Other recording proces~es than overwriting In the foregoing embodiments, the magnetooptical 20 recording medium of the present invention has been explained in relation to an overwriting process, but said medium can also be employed in the conventional recording process consisting of recording, erasure and re-recording. In such process there can be consi-25 dered a case in which the magnetizations of twomagnetic layers are in a mutually stable state in the recorded area, and another case in which said 16~1 134~0~1~

1 magnetizations are in a mutually stable state in the unrecorded area. These two cases will be separately explained in the following.
(i) Case of stable state in unrecorded area:
Recording is conducted on a magnetooptical r-cording medium of a structure as shown in Fig. 3, with an apparatus as shown in Fig. 6. The recording process is illustrated in Fig. 39, wherein it i5 assumed that the magnetizations of the magnetic layers 2, 3 are stable wh-n they are oriented in a same direction (parallel).
At first the magnetizations of the disk 9 are oriented upwards a~ shown by 80a in Fig. 39. An area to be recorded is locally heated, by a laser beam of a recording power from the record/reproducing head 5, to a temperature close to the Curie point of the ~econd magnetic layer 3. Simultaneously there is applied a bias ~magnetic field (downward in Fig. 39) of a magnitude which is necessary, or, preferably minimum necessary as will be explained later, for inverting the magnetization of the second magnetic layer 3. Thus, following the magnetic inversion of the second magnetic layer 3, the magnetization of the first magnetic layer 2 is also oriented in a stable direction (same direction in the present case) with respect to the magnetization of the second magnetic layer 3. In this manner a record bit 80b shown in Fig. 39 is formed.

162 l~ l qi~ ~

1 Upon rotation of the magnetooptical disk 9, said bit 80b passes through the pO5 ition of the magnetic field generating unit 8, of which field intensity is selected between the coercive forces HL
and HH of the two magnetic layers and of which field is oriented in a direction to invert the magnetization of the second magnetic layer in the bit 80b in Fig. 39, whereby the second magnetic layer 3 is magnetized along the direction of the field from said unit 8 as shown by 10 80c, while the first magnetic layer 2 retains the magnetization state 80b. In this manner the magnetiza-tions of the two magnetic layers are in a mutually opposite or antiparallel state, and ~uch state is utilized as the final record.

The record bit 80c can be erased in the following mAnner. The magnetooptical disk is rotated by a spindle motor, and the record bit, in passing the record/
repro~ducing head 5, is irradiated with a laser beam of an erasing power, whereby the irradiated area is 20 heated close to the Curie point of the first magnetic layer 2. Since the second magnetic layer 3 has a coercive force enough for retaining the bit in stable manner at 6uch temperature, the record bit is erased as shown by 80d in Fig. 39 through a suitable selection 25 of the bia~ magnetic field.
The suitable selection of the bias field has the following meaning.

16~ 13~00~?~

1 In this erasing process, the bias field is essentially unnecessary because the first magnetic layer 2 receives an exchange force to orient it6 magnetization in a stable direction (same direction in S this case) with re6pect to the magnetization of the second magnetic layer 3. However, as explained before, the bias field is provided, for the recording process, in a direction to assist the magnetic inversion of the second magnetic layer 3. It is convenient if said 10 bias field selected for the recording process can be maintained also in the erasing process with same magnitude and direction. Based on such consideration, the suitable selection of the bias field is achieved by selecting an intensity and a direction enabling the 15 recording process and not hindering the erasing process, preferably an intensity and a direction which are minimum necessary for the recording process, and such bias -field is then maintained also in the erasing process.

However it is also possible to remove the bias field at the erasing, since it is essentially unnecessary as explained before. Also it is not necessary to maintain the bias field in the same direction and magnitude in the recording and erasing processes, if the record/reproducing head is capable of varying the bias field according to the recording and reproducing operations. Since the recording and erasing operations 1 ~ 4 ~ ) 0 .1 ~

1 basically require mutually opposite bias fields, and such variation of the bias field enables recording are erasure at a higher spe-d.
Though Fig. 39 show~ the principle of recording and erasure in case the magnetizations of the first and second magnetic layers ~re stable when they are oriented in a same direction, a substantially same principle is applicable also if said magnetizations are stable when they are oriented in mutually opposite or 10 antiparallel directions. Such principle is illustrated in Fig. 41, wherein the states of magnetization 81a -81d respectively correspond to 80a - 80d shown in Fig. 39.
Example 36 A pregrooved and preformatted polycarbonate 15 disk substrate was set in a sputtering apparatus with ternary targets, and was rotated at a distan-ce of 10 cm from the targets.
A ZnS protective layer of 1000 A in thickness wa6 formed by sputtering, in argon gas from a first 20 target with a sputtering speed of 100 A and a sputtering pressure of 5 x 10 Torr. Then a TbFe alloy was sputtered, in argon gas, from a second target with a sputtering speed of 100 A and a sputtering pres6ure of 5 x 10 Torr to obtain a first magnetic layer of 25 Tbl8Fe82 with a thickness of 500 A, TL = ca. 140~C
and HH = ca. 10 KOe.
Then a TbFeCo alloy was sputtered, in argon gas, ~, 1 with a sputtering pressure 5 x 10 3 Torr to obtain a second magnetic layer of Tb23Fe73Co4 with a thickness of 500 A, TL = ca. 180~C and HL = ca. 1 KOe.
Subseguently a ZnS protective layer of 3000 A
5 in thickness was formed by sputtering, in argon gas, from the first target with a fiputtering speed of 100 A/min. and a sputtering pressure of 5 x 10 Torr.
After said layer formations, the above-mentioned substrate was adhered to a polycarbonate plate with 10 hot-melt adhesive material to obtain a magnetooptical disk.
Said disk was mounted on a record/reproducing apparatus, and was made to pass through a magnetic field generating unit of a field intensity of 2.5 KOe, 15 with a linear speed of ca. 15m/sec., and the recording and erasing were conducted with a laser beam of a wavelength of 830 nm, focused to ca. 1 ~m. The recording was conducted with a laser beam of 5 mW, modulated with a duty ratio of 50% and a freguency of 4 MHz.

The bias field was ca. 150 Oe, in a direction to invert the magnetization of the second magnetic layer. Intensity and direction of the bias field were maintained same, both at recording and at erasing.
The erasure was conducted by irradiation with 25 a continuous laser beam of various powers.
The reproduction was then conducted by irradiation with a laser beam of 1 mW. When the 166 131~d~

1 erasing laser power was equal to or larger than 5 mW, the remaining unerased signal was less than 10 dB at the reproduction and was saturated a~ shown in Fig. 40, and this level corresponds to an almost complete erasure.
Reference Examples 1, 2 A sample of Reference Example 1 was prepared in the same manner and structure as in the Example 36, except that a ZnS intermediate layer of 100 A was formed 10 between the first and second magnetic layers, which, in this case were magnetostatiaally coupled.
Also a sample of Reference Example 2 was prepared by forming the fir~t magnetic layer in a thickness of 1000 A only, instead of forming the first 15 and second magnetic layers, in a si~; l~r process as in the Example 36.
These samples were tested in recording and erasi~g in the same manner as in the Example 36. As shown in Fig. 40, the erasure required a power of ca. 7 20 mW.
It was confirmed that, in the Example 36 and in the Reference Examples 1 and 2, a signal with a C/N ratio of 45 - 50 dB could be recorded with a laser power of 5 mW.

(ii) Case of stable state in recorded area:
Recording i6 conducted on a magnetooptical recording medium of a structure as shown in Fig. 3, 167 1~100:~8 1 with an apparatus as shown in Fig. 6. The recording process is illustrated in Fig. 42, wherein it is assumed that the magnetizations of the magnetic layers 2, 3 are stable when they are oriented in a same (parallel) direction.
At first the magnetizations of the disk 9 are oriented as shown by 82a in Fig. 42. An area to be recorded is locally heated, by a la6er beam of a recording power from the reaord/reproducing head 5, 10 to a temperature close to the Curie point of the second magnetic layer 3. Simultaneously there is applied a bias magnetic field~upward in Fig. 42) of a magnitude which is necessary, or, preferably minimum necessary as will be explained later, for inverting the magnetiza-15 tion of the second magnetic~ layer 3. Thus, following themagnetic inversion of the second magnetic layer 3, the magnetization of the first magnetic layer 2 is also oriented in a stable direction (same direction in the present case) with respect to the magnetization of the second magnetic layer 3. In this manner there is formed a record bit as shown by 82b in Fig. 42. The bias field in this state is however not necessary, since, in this medium, the magnetizations are stable when they are oriented in a same direction.

Said bit can be erased by passing the disk 9 again under the record/reproducing head 5. Said area is locally heated, by a laser beam of an erasing ,~ , , 16~

1 power, close to the Curie point of the second magnetic layer 3. Simultaneously there is applied a bias magnetic field (downward in Fig. 42) of a magnitude which is necessary, preferably minimum necessary as 5 will be explained later, for inverting the magnetiza-tion of the second magnetic layer 3. Thus, following the magnetic inversion of the second magnetic layer 3, the magnetization of the fir~t magnetic layer 2 is oriented in a stable direction (same direction in 10 this case) with respect to the magnetization of the second magnetic layer 3.
Then the disk 9 is further rotated whereby the area 82c passes through the position of the magnetic field generating unit 8, of which field 15 intensity is selected at a value between the coercive forces HL, HH of the magnetic layers and of which field is oriented in a direction to invert the magnetization of the second magnetic layer 3 in the bit 83d, whereby the second magnetic layer 3 is magnetized 20 according to the direction of the magnetic field from the unit 8, while the first magnetic layer 2 retains the magnetization of the state 82c. In this manner the magnetic layers show mutually opposite magnetizations, thus returning to the state 82a. The erasure is 25 achieved in this manner.
Though Fig. 42 shows the principle of recording and erasure in case the magnetizations of the first 1 6 9 ~ o j ~
., 1 and second magnetic layers 2, 3 are stable when they are oriented in a same direction, a substantially same principle is applicable also if said magnetizations are stable when they are oriented in mutually opposite or antiparallel directions. Such principle is illustra-ted in Fig. 43, wherein the states of magnetization 83a - 83d respectively correspond to 82a - 82d in Fig.
42.
Ex~mple 37 A pregrooved and preformatted polycarbonate disk substrate was ~et in a sputter$ng apparatu~ with ternary targets, and was rotated at a distance of 10 cm from the targets.
A ZnS protective layer of 1000 A in thickness 15 was formed by sputtering, in argon gas, from a first target with a sputtering speed of 100 A/min. and a sputtering pres~rue of 5 x 10 Torr. Then a TbFe alloylwas sputtered, in argon gas, from a second target with a sputtering speed of 100 A/min. and a 20 sputtering pressure of 5 x 10 Torr to obtain a first magnetic layer of Tbl8Fe82 with a thickness of 300 A, TL = ca. 140~C and HH = ca. 10 KOe.
Then a TbFeCo alloy was sputtered, in argon gas, with a sputtering pressure of 5 x 10 3 Torr to 25 obtain a second magnetic layer of Tb23Fe63Col4 with a thickness of 500 A, TH = ca. 200~C and HL = ca. 1 ~Oe.
Subsequently a ZnS protective layer of 3000 A

1 in thickness was formed by sputtering, in argon gas, from the first target with a sputtering speed of 100 A/min. and a sputtering pressure of 5 x 10 3 Torr.
After said layer formations, the above-mentioned substrate was adhered to a polycarbonate plate with hot-melt adhesive material to obtain a magnetooptical disk.
Said disk was mounted on a record/reproducing apparatus, and was made to pass through, with a linear 10 speed of ca. 15 m/sec., a magnetic field generating unit of a field intensity of 2.5 KOe, and the recording and erasing were conducted with a laser beam of a wavelength of 830 nm focused to ca. 1 pm.
The recording was conducted with the laser 15 beam modulated with a duty ratio of 50% and a frequency of 4 MHz, and with various powers. The bias field was not used.
The erasure was conducted by irradiation of a continuous laser beam of 6 mW, in the presence of a 20 bias magnetic field of ca. 150 Oe in a direction to invert the magnetization of the second magnetic layer.
Reproduction was conducted with a laser beam of 1 mW. It was found that the recording was pos6ible with a power of ca. 3mW, as shown in Fig. 44.

25 Example 38 For achieving a recording with a higher speed than in the Example 37, a bias magnetic field of 150 Oe 17l ~34~ -1 wa~ applied in a direction, in which the magnetization of the fir~t magnetic layer was oriented at the recording. In the present example the direction of the magnetic field was inverted at the erasing, because the direction of bias field at recording is opposite to that at erasing. The necessary recording power was ca. 2.6 mW.
Example 39 Recording was conducted with a bias magnetic 10 field, which was maintained constant in direction and in intensity, both for the recording and erasing. The magnitude of the bias field enabling erasure and not hindering recording was ca. 200 Oe.
Reference Examples 3, 4 A sample of the Reference Example 3 was prepared in the same manner as in the Example 39, except that a ZnS intermediate layer of a thickness of l00 A was formed between the first and second magnetic layers. In this sample said layers were magneto-20 statically coupled.
Also a sample of the Reference Example 4 was prepared in the same manner, but by forming the first magnetic layer in a thickness of 800 A, instead of forming the first and second layers.

Recording and erasing were conducted on these samples, in the same manner as explained before.
The power required for recording was ca.4 mW as shown 1 7 ~ ~ 3 ~1 ~ O . i 3 1 in Fig. 44.
Erasure was proved with a laser power of 6 mW, on the samples of the Example 37 and the Reference Examples 3 and 4.
5 Example 40 Recording was conducted in the same mAnner as in the Example 37.
Erasing was conducted with the following method, which is suitable for the erasure of the entire 10 surface of the disk or the erasure of plural tracks without excessive precision. More specifically, the erasing was achieved by another magnetic field genera-ting unit, which was positioned between the recording head S and the magnetic field generating unit 8 and 15 which is capable of generating a magnetic field of a direction opposite to that of the field generated by said unit 8 and of a magnitude enough for inverting the magnetization of the first magnetic layer. In this manner it is no longer necessary to effect the erasure 20 with the record/reproducing head.
Improvement of the apparatus Fig. 45 is a schematic view of another embodiment of the magnetooptical recording apparatus of the present invention, wherein same components as those in 25 Fig. 8 are represented by same numbers and will not be explained further. In the present embodiment, the first magnetic field generating means 24 is replaced by a ,. , ~, , , ~ , ... . . .. . ...

1 leading field of a driving magnet of an actuator 84 for moving the objective lens 19 of the optical head 14.
Also the ~econd magnetic field generating means 25 ifi positioned at the same side as the optical head 14, with respect to the disk 11. It is therefore rendered po6sible to reduce the number of component parts and to compactize the apparatus.
Fig. 46 is a partially cut-off plan view showing still another embodiment, in which same com-10 ponents as those in Fig. 45 are represented by samenumbers and will not be explained further. In this embodiment the second magnetic field generating means 85 is mounted on the optical head 14, thereby achieving further compactization of the apparatus.

The present invention is not limited to the foregoing embodiments, but i6 subject to various modifications within the scope and spirit of the appended claims.

Claims (32)

1. A magnetooptical recording medium provided with a first magnetic layer and a second magnetic layer having a higher Curie point and a lower coercive force than those of said first magnetic layer and being exchange coupled with said first magnetic layer; which comprises satisfying a relation:

wherein H H is the coercive force of the first magnetic layer;
H L is the coercive force of the second magnetic layer; M s is the saturation magnetization of the second magnetic layer; h is the thickness of the second magnetic layer; and .sigma. w is the magnetic wall energy between the first and second magnetic layers.

2. A magnetooptical recording medium according to claim 1, wherein said first and second magnetic layers are composed of alloys selected from GdCo, GdFe, TbFe, DyFe, GdTbFe, TbDyFe, GdFeCo, TbFeCo and GdTbCo.

3. A magnetooptical recording medium according to claim 1, wherein said first magnetic layer has a Curie point in a range from 70 to 180°C, and said second magnetic layer has a Curie point in a range from 150 to 400°C.

4. A magnetooptical recording medium according to claim 1, wherein said first magnetic layer has a coercive force in a range from 3 to 10 KOe, and said second magnetic layer has a coercive force in a range from 0.5 to 2 KOe.

5. A magnetooptical recording medium according to claim 1, further comprising a protective layer at least on one side of said first and second magnetic layers.

6. A magnetooptical recording medium provided with a first magnetic layer and a second magnetic layer having a higher Curie point and a lower coercive force than those of said first magnetic layer and being exchange coupled with said first magnetic layer, wherein there is provided between said first magnetic and said second magnetic layer an adjusting layer for adjusting a magnetic wall energy .sigma. w between said first and second magnetic layers in such a manner that the magnetic wall energy .sigma. w satisfies a relation:

where H H is the coercive force of the first magnetic layer;
H L is the coercive force of the second magnetic layer; M s is the saturation magnetization of the second magnetic layer;
and h is the thickness of said second magnetic layer.

7. A magnetooptical recording medium according to claim 6, wherein said adjusting layer is composed of a material selected from Ti, Cr, Al, Ni, Fe, Co, rare earth elements and transition metals and has a thickness in a range from 5 to 50 .ANG..

8. A magnetooptical recording medium according to Claim 6, wherein said adjusting layer is composed of a magnetic material having a coercive force H and the following relationship is satisfied:
H H>H L>H

9. A magnetooptical recording medium according to Claim 8, wherein the following relationship is satisfied:

T2~T>T1 wherein T1, T2 and T are respectively Curie temperatures of said first magnetic layer, said second magnetic layer and said adjusting layer.

10. A magnetooptical recording medium according to Claim 9, wherein said first magnetic layer, second magnetic layer and adjusting layer are composed of alloys of rare earth element and transition metal, in which the first magnetic layer has a composition rich in the transition metal compared with the compensation composition, said second magnetic layer and adjusting layer have compositions both rich in the rare earth element compared with the compensation compositions.

11. A magnetooptical recording medium according to Claim 9, wherein the coercive force H of said adjusting layer is in the range of 0.1 to 1 KOe and the Curie temperature T of said adjusting layer is in the range of 90 to 400°C.

12. A magnetooptical recording medium according to Claim 8, further comprising a third magnetic layer formed on the opposite side of the first magnetic layer to the side faces to the adjusting layer and exchange-coupled to the first magnetic layer, wherein the following relationship is satisfied:
(I) as for the Curie temperatures of the magnetic layers:
T~,T2>TI,T
(II) as for the coercive forces of the magnetic layers:

H H>H L>H3,H

(III) as for the saturation magnetizations, thicknesses, coercive forces and magnetic wall energies of the magnetic layers:

, , , , , , wherein T1, T2 and T3 are respectively Curie temperatures of the first, second and third magnetic layers, T is a Curie temperature of the adjusting layer, H3 is a coercive force of the third magnetic layer, h1, h2 and h3 are respectively thicknesses of the first, second and third magnetic layers, h4 is a thickness of the adjusting layer, .sigma.w12, .sigma.w23 and .sigma.w34 are magnetic wall energies respectively for the first and third magnetic layers, the first magnetic layer and the adjusting layer, and the adjusting layer and the second magnetic layer, Ms1, Ms3 and M s4 are respectively saturation magnetization of the first magnetic layer, the third magnetic layer and the adjusting layer.

13. A magnetooptical recording medium according to Claim 12, wherein the following relationship is satisfied:

h1+h3 ~ 250.ANG.
h1+h2+h3+h4 ~ 600.ANG..

14. A magnetooptical recording medium according to Claim 1, further comprising a third magnetic layer formed on the opposite side of the first magnetic layer to the side faces to the second magnetic layer and exchange-coupled to the first magnetic layer, wherein the following relationship is satisfied:

wherein H3 is a coercive force of the third magnetic layer, h3 is a thickness of the third magnetic layer, .sigma.W12 is a magnetic wall energy for the first and third magnetic layers, M~3 is a saturation magnetization of the third magnetic layer.

15. A magnetooptical recording medium according to Claim 14, wherein the Curie temperatures and coercive forces of said second and third magnetic layers satisfy relations:

T3 ~ T2 and H3 ~ H L

wherein T2 and T3 are respectively Curie temperatures of said second and third magnetic layers, H3 is a coercive force of the third magnetic layer.

16. A magnetooptical recording medium according to Claim 14, wherein the coercive force H3 of said third magnetic layer is in the range of 0.1 to 1 KOe and the Curie temperature T3 of said third magnetic layer is in the range of 150 to 400°C.

17. A magnetooptical recording medium according to Claim 14, wherein said first, second and third magnetic layers are composed of alloys of rare earth element and transition metal, wherein the first and third magnetic layers have compositions both rich in the transition metal compared with the compensation composition, while the second magnetic layer has a composition rich in the rare earth element compared with the compensation composition.

18. A magnetooptical recording medium provided with a first magnetic layer with a Curie point T1 and a coercive force H1, and a second magnetic layer with a Curie point T2 and a coercive force H2 and exchange coupled with said first magnetic layer, which comprises satisfying following conditions:

at room temperature wherein M~2 is the saturation magnetization of the second magnetic layer, .sigma. w is the magnetic wall energy between two magnetic layers, and h2 is the thickness of the second magnetic layer; and (B) T1comp < room temp. < T2comp < T1 ~ T2 wherein T1comp is the compensation temperature of the first magnetic layer; and T2com is the compensation temperature of the second magnetic layer.

19. A magnetooptical recording medium according to Claim 18, wherein said second magnetic layer has a compensation temperature T2comp in a range from 50 to 180°C.

20. A magnetooptical recording medium according to Claim 18, wherein said first and second magnetic layers are composed of materials selected from GdCo, GdFe, TbFe, DyFe, GdTbFe, TbDyFe, GdFeCo, TbFeCo, and GdTbCo.

21. A process for recording information on a magnetooptical recording medium according to any one of Claims 1 to 20, said process comprising steps for:
(a) orienting the direction of magnetization of the second magnetic layer in a predetermined direction while retaining the direction of magnetization of the first magnetic layer; and (b) selectively effecting either a first type of recording by irradiating the medium with a light beam having a power sufficient to heat the medium close to the Curie temperature of the first magnetic layer, thereby orienting the direction of magnetization of the first magnetic layer in a direction that is stable with respect to the direction of magnetization of the second magnetic layer, or a second type of recording by irradiating the medium with a light beam having a power sufficient to heat the medium close to the Curie temperature of the second magnetic layer while applying a bias magnetic field, thereby inverting the direction of magnetization of the second magnetic layer and simultaneously orienting the direction of magnetization of the first magnetic layer in a direction that is stable with respect to the direction of magnetization of the second magnetic layer, according to an information signal.

22. A magnetooptical recording apparatus which is capable of overwriting information on a magnetooptical recording medium according to any one of Claims 1 to 20, said apparatus comprising:
first magnetic field generating means for applying a magnetic field of a magnitude enough for magnetizing said second magnetic layer in one direction but incapable of inverting the direction of the magnetization of said first magnetic layer;
means for irradiating the magnetooptic recording medium that has passed said first magnetic field generating means, with a light beam which is intensity modulated between two non-zero values according to said information; and second magnetic field generating means for applying a bias magnetic field for orienting the direction of magnetization of said second magnetic layer at a position of light beam irradiation in a predetermined direction in a case where said irradiating means irradiates said medium with the light beam of intensity of a higher value.

23. A magnetooptical recording apparatus which is capable of overwriting information on a magnetooptical recording medium according to any one of Claims 1 to 20, said apparatus comprising:
first magnetic field generating means for applying a magnetic field of a magnitude enough for magnetizing said second magnetic layer in one direction but incapable of inverting the direction of the magnetization of said first magnetic layer;
means for irradiating the magnetooptic recording medium that has passed said first magnetic field generating means, with a light beam which is intensity modulated between two non-zero values according to said information; and second magnetic field generating means for applying a bias magnetic field for orienting the direction of magnetization of said second magnetic layer at a position of light beam irradiation in a predetermined direction in a case where said irradiating means irradiates said medium with the light beam of intensity of a higher value;
wherein said light beam irradiation means is an optical head comprising a light source and an optical system for focusing a light beam from said light source onto said medium.

24. A magnetooptical recording apparatus which is capable of overwriting information on a magnetooptical recording medium according to any one of Claims 1 to 20, said apparatus comprising:
first magnetic field generating means for applying a magnetic field of a magnitude enough for magnetizing said second magnetic layer in one direction but incapable of inverting the direction of the magnetization of said first magnetic layer;
means for irradiating the magnetooptic recording medium that has passed said first magnetic field generating means, with a light beam which is intensity modulated between two non-zero values according to said information; and second magnetic field generating means for applying a bias magnetic field for orienting the direction of magnetization of said second magnetic layer at a position of light beam irradiation in a predetermined direction in a case where said irradiating means irradiates said medium with the light beam of intensity of a higher value;
wherein said first and second magnetic field generating means are composed of magnets positioned close to said medium.

25. A method for producing a magnetooptical recording medium according to Claim 1, said method comprising the steps of:
(a) arranging a transparent substrate in a vacuum apparatus;
(b) forming a first magnetic layer on the transparent substrate by sputtering or deposition;
(c) leaving the first magnetic layer formed in the step (b) for five minutes or more under a condition in which a pressure of residual gas or inert gas in the apparatus is 7 X 10-7 Torr. or more; and (d) forming by sputtering or deposition, on the first magnetic layer, a second magnetic layer having a higher Curie point and a lower coercive force at room temperature than those of the first magnetic layer.

26. A method according to Claim 25, wherein said inert gas consists of one selected from the group consisting of Ar, He and Ne.

27. A method for producing a magnetooptical recording medium, according to Claim 1, said method comprising the steps of:
(a) arranging a transparent substrate in a vacuum apparatus;
(b) forming a first magnetic layer on the transparent substrate by sputtering or deposition;
(c) supplying into the apparatus gas consisting of an element which reacts on a constituent element of the first or second magnetic layers or which can be adsorbed to the constituent element; and (d) forming by sputtering or deposition, on the first magnetic layer, the second magnetic layer having a higher Curie point and a lower coercive force at room temperature than those of the first magnetic layer.

28. A method according to Claim 27, wherein said gas consists of one of the group consisting of H2O, O2, H2, N2, H2S, CS2 and CH4.

29. A method according to Claim 27, wherein a partial pressure of the supplied gas in the apparatus is 2 x 10-6 Torr. or more.

30. A method for producing a magnetooptical recording medium, according to Claim 1, said method comprising steps of:
(a) arranging a transparent substrate in a vacuum apparatus;
(b) forming a first magnetic layer on the transparent substrate by sputtering or deposition;
(c) effecting a plasma discharge in an atmosphere of inert gas or gas consisting of an element which reacts on a constituent element of the first or second magnetic layers or which can be adsorbed to the constituent element;

(d) forming by sputtering or deposition, on the first magnetic layer, the second magnetic layer having a higher Curie point and a lower coercive force at room temperature than those of the first magnetic layer.

31. A method according to Claim 30, wherein said reactive or adsorbent gas consists of one selected from the group consisting of H2O, O2, H2, N2, H2S, CS2 and CH4.

32. A method according to Claim 30, wherein said inert gas consists of one selected from the group consisting of Ar, He and Ne.