JPH09314839A - Ink jet recording head - Google Patents

Abstract

PROBLEM TO BE SOLVED: To secure a basically necessary performance and optimize the height of an ink flow path and the nozzle length and lower the production cost by a method wherein the height of an ink flow path is determined to be a specific dimension or less and the shape of an ink discharge nozzle is generally cylindrical and the length is determined to be a specific dimension or less. SOLUTION: A simulation relating to a head of a construction such that the image density is not changed with temperature is effected, and the height of an ink barrier (ink flow path) is lowered without marring the superior characteristics of the head and the thickness of an orifis plate (the thickness of a cylindrical type nozzle) is thinned then an optimization is effected. When a barrier height is 10μm, a retention ink quantity drawn by a discharge ink is small, and the effect on the whole discharge ink quantity is low. Therefore, the barrier height is determined to be 15μm or less and the nozzle length is determined to be 35μm or less. Whereby, the thickness of the barrier and the orifis plate can be lowered and greatly contribute to side and cost reduction.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an ink jet recording head of a type in which ink droplets are ejected toward a recording medium by pulse heating.

[0002]

2. Description of the Related Art Two methods have been put into practical use for an ink ejection head used in a thermal ink jet printer. One is that the heating resistor formed on one wall surface (substrate) of the ink liquid path and the ink ejection direction are parallel to each other (JP-A-54-161935 and JP-A-55-55). 27
No. 281, Japanese Patent Laid-Open No. 27282: hereinafter referred to as side shooter type), and the other one is vertical (see Japanese Laid-Open Patent No. 54-51837: hereinafter referred to as top shooter type). . In both cases, a portion of the ink is rapidly vaporized by pulse heating, and the ink droplets are ejected from the orifice by its rapid expansion and contraction. The basic structure of the heating resistor is a thin film resistor It is also the same in that it is covered with a protective layer of Nikkei Mechanical (December 28, 1992, p. 58, and Hewlett-Pa
ckard-Journal, Aug. 1988).

It is well known that the image density changes depending on the head temperature as a characteristic common to the ink ejection heads of these two types that have been put to practical use.

The reason why the image density depends on the head temperature is as follows.
This is related to the fact that the viscosity of ink greatly changes with temperature. Therefore, it is essential to adopt various complicated control means in an actual printer as a measure to prevent the image density from changing (for example, Electrophotographic Society Journal, Vol. 33 (No. 2) (1994) 177).

Further, since the thin film resistor is obliged to be covered with a multi-layered thick protective layer, the applied energy required for ejecting ink is as large as 15 to 30 μJ / pulse, which causes the head temperature to rise rapidly. It is the cause. Furthermore,
This multi-layered thick protective layer keeps the maximum temperature rising rate of the heating surface in contact with the ink to 7 × 10 ⁷ ° C / s or less, and is likely to be an obstacle in that stable nucleate boiling that is indispensable for stable ink ejection is generated. Is known (see Proceedings of the 25th Japan Heat Transfer Symposium (1988-6) P177).

With respect to these problems, the present inventor has developed the following drastic improvement method. That is, even if the head temperature changes, the image density does not change, and the ink droplets have a short tail structure (Japanese Patent Application Laid-Open No. 07-227967, Japanese Patent Application No. 06-156949, Japanese Patent Application No. 07-43968). No.), about 100Å
A thin-film heating resistor (see Japanese Patent Application No. 07-340486) having sufficient reliability with only an ultra-thin self-oxidation film, and the energy input for ink ejection is greatly reduced to about 3 μJ / pulse. For more stable ink ejection, the optimum heating rate of the heater surface in contact with the ink is 1 × 10 ^{8 to} 5 ×
It was clarified that it must be in the range of 10 ⁸ ° C / s (Japanese Patent Application No. 07-285650), and it was clarified that this temperature rising condition can be easily realized by using the above thin film heating resistor.

[0007]

Regarding the top shooter type head, the head structure in which the image density does not fluctuate even if the head temperature changes is as follows: (1) Perpendicular to the heating resistor at the bottom of the ink discharge nozzle The projected image overlaps the heating resistor within ± 5 μm. (2) The height of the ink liquid path is within 30 μm. (3) The shape of the ink discharge nozzle is almost cylindrical and its length is 80 μm. That is within.

Further, the inventor of the present invention has developed a method for manufacturing such a head in a unit of tens of thousands of nozzles or more, and applied for a patent (Japanese Patent Application No. 07-135185, Japanese Patent Application No. 07-288877). , Japanese Patent Application No. 07-320446, Japanese Patent Application 07-334802, Japanese Patent Application 07-340486
No., etc.).

The greatest feature of this manufacturing method is that the head is manufactured only on the same thin film process as the semiconductor process on the Si wafer in which the driver circuit is formed.

However, for the thin film process, the above-described ink liquid path height and nozzle length are as low as possible,
The shorter the length, the easier the manufacturing, and in that case, the manufacturing cost of the head can be further reduced.

An object of the present invention is to optimize the height of an ink liquid path and the nozzle length while ensuring excellent characteristics and basically required performance.

[0012]

Means for Solving the Problems The above-described problems are solved by an ink ejection nozzle for ejecting ink in a direction perpendicular or almost perpendicular to a thin film heating resistor, and an ink ejection nozzle for supplying ink to this ink ejection nozzle. And an ink liquid path communicating with the ink discharge path, wherein a vertical projection image of the bottom of the ink discharge nozzle onto the heat generating resistor overlaps with the heat generating resistor within ± 5 μm. Has a height of 15 μm or less, the ink discharge nozzle has a substantially cylindrical shape, and its length is 35 μm.
It is achieved by setting the distance to be within m.

[0013]

BEST MODE FOR CARRYING OUT THE INVENTION As described above, in order to optimize the shapes of ink liquid paths and nozzles, a great number of types of heads must be prototyped and evaluated under various conditions. Therefore, we adopted a method that efficiently obtains this by computer simulation and confirms it by experiments.

The simulation program used is Flow
Flow-3D (registered trademark) manufactured by Science (USA), which can three-dimensionally solve the ejection of ink droplets from the head. However, such a fluid analysis program cannot handle phase changes such as nucleate boiling and condensation due to adiabatic expansion of water vapor bubbles.

Therefore, the nucleate boiling is replaced with a high-pressure air film on the heater as an initial condition, and the vacuumization of the foamed bubbles by condensation is set to a boundary condition of an external pressure of 2 atm, whereby a negative pressure of about 1 atm is simulated. The pressure was obtained equivalently. Then, the initial conditions (pressure and volume of the high-pressure air film) at which the experimental results and the simulation results by the stroboscopic observation of the bubble generation, growth, and disappearance in the open pool boiling match are found, and these initial conditions are the heat transfer analysis of pulse heating. It was confirmed that it is in good agreement with the initial appearance of nucleate boiling obtained from. In other words, the simulation of Flow-3D that solves under the initial conditions and boundary conditions determined in this way allows the ink movement to be performed with fairly good accuracy even for the head that ejects ink by pulse heating under the same conditions. It can be expected to reproduce it (we will explain later that it actually agrees well with the experimental result).

Here, the appearance of nucleate boiling at an early stage, which is obtained from the heat transfer analysis of pulse heating, will be briefly described.

The thin film heating resistor we use has a thickness of about 1
About 1000Å T with 00Å self-oxidizing insulation film
a-Si-O ternary alloy thin film (see Japanese Patent Application No. 07-340486)
And is formed on the SiO ₂ heat insulating layer having a thickness of about 2 μm formed on the Si substrate. For example,
Pulse width 1μs, applied power 2.4W / 50μm □ By applying pulse current to the heater, the ink in contact with this is heated to around 320 ° C, and caviar nucleate boiling is generated in the ink to eject the ink. (Japanese Patent Application No. 07-285650). The temperature at which the caviar nucleate boiling starts is around 300 to 310 ° C., and the boiling pressure (saturated vapor pressure) at this time is 86 to 100 atm in the case of pure water, and the aqueous solution to which a small amount of alcohol as a foaming aid is added. In the case of ink, it is estimated to be a little higher, so it is treated as 100 atm here.

Since the heating time is as short as about 1 μs, the ink flow during this period can be ignored. Therefore, the one-dimensional heat transfer analysis can be used to almost accurately solve how the heat flux input to the heater heats the ink and the SiO ₂ layer (for example, the Japan Society of Mechanical Engineers: Heat Transfer Engineering Material 4th Edition, Maruzen). (1989)).

For example, if it is assumed that nucleate boiling does not occur in the temperature distribution in the ink and the SiO ₂ layer after 1 μs of pulse heating, it is obtained as shown in FIG. That is, the area heated by pulse heating in a short time of 1 μs is 1 μs on the ink side.
It can be understood that the one-dimensional analysis is appropriate for a heater having a small area of 50 μm and a depth of 2 μm on the SiO ₂ layer side.

On the other hand, on the surface of the heater at the time of FIG. 1, the caviar-like nucleate boiling is already saturated in reality, and therefore the nucleate boiling is 305 to 317 ° C. as shown in FIG.
It is considered that the ink (average temperature: 311 ° C.) heated to is involved in the latent heat (the same amount of heat that heats the ink at 20 ° C. to 310 ° C.) is supplied from the adjacent ink and heater at 305 ° C. or less. . Then, the ink heated to 305 to 317 ° C. (average temperature: 311 ° C.) is found to be the ink having a thickness of about 220 Å in contact with the heater, so this is 311 ° C., 100 atm, and volume of 0. It will change to saturated steam of 4 μm × (heater area) (this is obtained from the saturation table of water). That is, if the ink is rapidly heated to about 320 ° C. in a short time of 1 μs with a heater equivalent to the state where the thin film resistor is in direct contact with the ink, the thickness on the heater is 1 μs after pulse heating. It is considered that a saturated water vapor film of 0.4 μm at 311 ° C. and 100 atm is generated. However, this saturated water vapor film undergoes adiabatic expansion rapidly and its volume is 10 to 2
When it is made 0 times (5-6μ thick), its saturated steam is -2.
It can be presumed that the bubbles are already in a vacuum state as they are cooled to 0 to −40 ° C. and condensed.

In the case of caviar nucleate boiling, since the heating rate is too fast, foaming begins on the entire area of the heater surface, and the latent heat consumed by the foaming cools the surroundings of the water vapor bubbles, so that the bubbles are small caviar-shaped. The density is rapidly increased as it is. That is, the thermal ink layer having a thickness of about 220 Å is about 0.4 μm in a short time of 0.2 to 0.3 μs.
Phase change to a high temperature and high pressure steam layer with a thickness of 5 to 6 μm
It can be inferred that the actual image of the caviar nucleate boiling is that the caviar bubble group in the vacuum state has a two-dimensionally expanded thickness.

Therefore, this nucleate boiling is 5 to 6 μm / 0.2.
The ink on the heater is pushed upward at an initial velocity of ~ 0.3 μs = 20 to 25 m / s, but the bubble group having a height of 5 to 6 μm and already in a vacuum state becomes a vacuum bubble having a height of 20 to 25 μm. It can also be understood that the growth up to is due to the rising inertial flow of ink.

The conclusions about the caviar nucleate boiling obtained from this heat transfer analysis are as follows.
It is given as 00 atm and thickness of 0.4 μm × (heater area). (2) After the bubble grows to a height of several μm, it becomes a vacuum state, but this makes the external pressure of 2 atm as a boundary condition. This means that an almost accurate simulation is possible.

The appearance and growth of bubbles in the open pool boiling solved under the initial conditions and the boundary conditions and the disappearance of the bubbles and the time thereof are completely in agreement with the experimental results.

An example of the simulation result of the open pool boiling is shown in FIG. The heater in this case is 4
Although the disk shape is 5 μmφ, the actual heater is equivalent to a 40 μm square heater.

Next, the result of simulation performed on a head having a structure in which the image density does not change with temperature will be described.

As described above, the purpose here is to reduce the height of the ink partition (ink liquid path) and to reduce the thickness of the orifice plate (the length of the cylindrical nozzle) without impairing the excellent characteristics of this head. Is to optimize it. From the preliminary examination in line with this purpose, for example, 400d
The basic structure of the pi (dot / inch) top shooter type is as follows.

That is, a thin film heating resistor having only an extremely thin self-oxidizing film of 40 μm □ is placed in the center of a heater chamber of 45 μm □, and a cylindrical nozzle of 45 μmφ and a length of 30 μm is arranged directly above this. 45 to the heater room from one direction
The structure is such that individual ink liquid paths each having a width of μm, a height of 10 μm, and a length of 30 μm are connected. Then, the ink liquid path has a density of 1 g / cm ³ , a viscosity of 2.5 cps, and a surface tension of 4
A standard ink having a dyne / cm of 5 and a contact angle of 60 ° is filled, and a simulation of ink ejection is performed under the initial conditions and boundary conditions determined by open pool boiling. The parameters changed in this simulation are the ink partition wall height (10 μm and 20 μm), the orifice plate thickness (20 μm, 30 μm, 40 μm), and the ink viscosity (1.3 cps, 2.5 cps, 6.0 cps).
And

First, under the condition that the maximum value of the height of the ink partition wall is within the maximum growth height (30 μm) of the bubbles, the combination of the basic structure head and the standard ink has two cases where only the height of the ink partition wall is 10 μm and 20 μm. Was examined.
In the simulation, air bubbles remain even after 6 μs, but as described above, it is actually in a vacuum state, and rebound does not occur in the head having this structure (see Japanese Patent Application No. 07-43968). Therefore, 6 of the simulation results after this including FIG.
A correction from that point of view has been added to the part after μs.

As can be seen from FIG. 3, when the partition wall height is 20 μm, the ink on the ejection side and the ink on the residual side are not completely separated by the air bubbles, and the timing at which the ink is cut by the ejected ink depends on the ink viscosity. It shows the possibility of change. On the other hand, the partition wall height is 1
In the case of 0 μm, it can be seen that the residual ink amount drawn out by the ejected ink is small, and even if the cutting timing slightly varies depending on the ink viscosity, the influence on the entire ejected ink amount is very small. of course,
There is additional supply of ink from the individual ink channels,
The flow path resistance is higher than that of the residual ink remaining in the heater chamber, and its effect can be ignored. In any case, these are proved by the experimental facts shown later.

From the above results, it is a necessary condition that the partition wall height be within 15 μm, and it can be estimated that the partition wall height of 7 to 10 μm is the optimum value for the nozzle of 400 to 800 dpi. It should be noted that if the partition wall height is made too low, the refilling time of ink is lengthened and the repetition frequency of ejection is lowered. Therefore, optimization of the shape of the individual ink flow path is necessary from this viewpoint, but it is omitted here.

FIG. 4 compares three cases in which the thickness of the orifice plate (nozzle length) is 20, 30, and 40 μm among the combinations of the basic structure head and the standard ink. From this result, it is a necessary condition that the nozzle length be within 35 μm, and even in the case of an extremely short nozzle of 10 μm, except that the amount of ejected ink is small,
It is estimated that it is possible to realize the ink ejection without causing the ink splash. And the nozzle length is 20 μm
Even in case of, 60 pl which is necessary and sufficient for printing 400 dpi
Therefore, the criterion for selecting the thickness of the orifice plate between 20 and 35 μm is as follows.
It is understood that the thickness may be determined from the viewpoint of production technology of uniformly laminating the orifice plate on the partition wall. It is easy to accurately form a cylindrical nozzle hole in the orifice plate of this thickness by dry etching, and the etching time is shortened to about 1/3 of the conventional one, which can greatly contribute to the improvement of productivity. FIG. 5 shows the results obtained by arranging the ink discharge speed obtained from this simulation by the ink depth on the heater (partition wall height + nozzle length). Simulation results for a nozzle length of 50 μm are also added here.

From FIG. 5, the ejection speed is 22.5 m / s when the depth of ink on the heater is zero.
This discharge velocity should correspond to the initial expansion velocity of the caviar nucleate boiling, and it can be seen that it agrees well with the predicted value of 20 to 25 m / s from the heat transfer analysis described above.

FIG. 6 shows only the ink viscosity of 1.3 cps in the combination of the basic structure head and the standard ink.
It is a simulation result when changing to 2.5 cps and 6.0 cps.

This change in viscosity corresponds to a case where the ink is changed from 5 ° C. to 60 ° C., and even in the case of a head using a conventional heater with a large input power, the change in temperature is only about half of this. That is.

FIG. 6 shows only the ejection state at the time of 9 μs after the pulse application, but the state before and after that is almost the same, so it is omitted.

In comparison with the conventional head in which the amount of ejected ink increases by about 50% when the temperature rises by 20 to 30 ° C., it is 50 to 6
The characteristics of the head of this structure, in which the amount of ejected ink does not change even when the temperature changes by 0 ° C., are well shown.

It can be seen from this that the ink viscosity does not affect the flight speed of the ejected ink.

FIG. 7 shows a stroboscopic image recording in which a head having a structure almost the same as the basic structure head is manufactured as a prototype, and the ink whose viscosity is adjusted by mixing glycerin in pure water is filled and ejected.

The difference from the basic structure head is that the thickness of the orifice plate is a little thick, 33 μm, and the size of the heating resistor is a little small, 36 to 37 μm □.

From the results of this experiment, the following facts can be confirmed. (1) Glycerin concentration 20% (ink viscosity 2.6 cps)
The ejection state up to 15 μs of the experimental result of 1) is in good agreement with the simulation result of the nozzle length of 30 μm in FIG. (2) The ejected ink amount does not change with ink having a glycerin concentration of 5% (ink viscosity 1.3 cps) to 50% (ink viscosity 6.0 cps). This is consistent with the simulation result shown in FIG. (3) However, the ejection speed is slower than the simulation result, and changes greatly depending on the glycerin concentration (ink viscosity).

This disagreement (3) is due to the fact that the amount of nucleate boiling of the ink is reduced and the ejection force is reduced by mixing glycerin. That is, as shown in FIG. 8, the ejection speed obtained from FIG. 7 linearly decreases in proportion to the glycerin concentration, and tends to become zero toward the glycerin concentration of 100%.

On the other hand, glycerin has a high boiling point at 1 atmosphere of 290 ° C., which is the caviar nucleate boiling temperature of pure water.
Even when heated to 00 to 310 ° C, the vapor pressure does not reach 2 atm. That is, glycerin dissolved in water is
The amount of nucleate boiling of water is reduced by the amount of dissolution, and the ejection force for ink is reduced in proportion to the concentration. The result of FIG. 7 shows that the amount of ejected ink does not change even when the ejection force becomes small in this way, which shows an excellent feature of the head of this structure.

On the other hand, in the case of the conventional head, the method of controlling the applied pulse to change the ejection force, thereby changing the amount of ejected ink to control the image density, is adopted. It has completely different characteristics from the structural head (for example, the Electrophotographic Society of Japan, 33 (2)
(1994) 177, Hewlett-Packard-Journal, Feb. (1994) 9).

It was found from these experiments that this difference was in the structure of the head of the present application, and the details thereof became clear from the thermal analysis simulation described in the present application. At the same time, it can be understood that even if a conventional heater having a thick protective layer is used, some fluctuation may occur, but a head in which the ejection amount does not change can be obtained. Needless to say, the print density does not change even if the head temperature is changed using the same ink.

The applied power of our head was as small as 1/5 to 1/10 of that of the head of the prior art, and it was not possible to raise the head temperature during printing and acquire data.
For this reason, experiments were conducted by variously adjusting the ink to which glycerin was added as described above.

Thus, both the thickness of the partition wall and the thickness of the orifice plate can be reduced to 1/2 to 1/3 in order to collectively manufacture 50,000 to 100,000 nozzles or more on one substrate by only the thin film process. It will be a great advantage. That is, the reduction in the number of steps, the reduction in processing time, and the improvement in yield due to these contributes to a significant reduction in the head manufacturing cost. Then, the cartridge equipped with this head becomes small in size and low in cost, and the recording apparatus using this cartridge has a great effect that the control system can be greatly reduced.

[0048]

According to the present invention, the thickness of the partition wall and the orifice plate of the head can be reduced by theoretically, numerically (simulating) and experimentally optimizing the head structure in which the image density does not change. It was found that both can be reduced to 1/2 to 1/3. These contribute to rationalization of head production and yield, and can also contribute greatly to downsizing and cost reduction of the cartridge and the recording apparatus in which the head is mounted.

[Brief description of drawings]

FIG. 1 is a graph showing temperature distributions in ink and a substrate after 1 μs of pulse heating.

FIG. 2 is a schematic diagram showing an example of a simulation result of open pool boiling.

FIG. 3 is a diagram showing a simulation result when the partition wall height is changed.

FIG. 4 is a diagram showing a simulation result when the thickness of the orifice plate is changed.

FIG. 5 is a graph showing the relationship between ink depth on a heater and ejection speed.

FIG. 6 is a diagram showing a simulation result when ink viscosity is changed.

FIG. 7 is a diagram showing a relationship between glycerin concentration in ink and ejected ink.

FIG. 8 is a graph showing the relationship between glycerin concentration and ink ejection speed.

Claims (2)

[Claims] 1. An ink ejection nozzle for ejecting ink in a direction perpendicular or almost perpendicular to a thin-film heat generating resistor, and an ink liquid passage communicating with the ink ejection nozzle for supplying ink to the ink ejection nozzle. And a vertical projection image of the bottom of the ink discharge nozzle onto the heating resistor overlaps the heating resistor within ± 5 μm, the height of the ink liquid path is within 15 μm. The ink discharge nozzle has a substantially cylindrical shape and its length is 35
An inkjet recording head characterized by being within μm. 2. The thin-film heating resistor comprises a Ta-Si-O ternary alloy thin-film resistor having a self-oxidizing film with a thickness of about 100Å, and the thin-film resistor is pulsed at 1 × 10 ⁸ ° C. 2. The ink jet recording head according to claim 1, wherein the surface temperature of the ink in contact with the ink is heated to about 320 [deg.] C. at a heating rate in the range of / s to 5 * 10 < ⁸ > C / s.