Chapter 111: The Glass Masters
Location: Shergill Optical Materials Division, Firozabad, Uttar Pradesh
Date: 5 March 1973 — 08:45 Hours
The furnace had been burning continuously since October 1971.
Seventeen months of uninterrupted operation at temperatures that would melt steel, fed by natural gas from the Bombay High fields, consuming thirteen tonnes of raw materials per day and producing optical glass that met specifications previously achievable only by German, Japanese, and American manufacturers.
Karan stood in the observation gallery above the main melting hall, looking down at furnace number three through heat-resistant glass that made the scene below shimmer slightly in the rising thermal currents.
Inside the furnace — a refractory-lined chamber three meters wide and two meters deep — molten glass glowed orange-white at 1450 degrees Celsius. Not the soda-lime glass of windows and bottles, but borosilicate optical glass, a precisely formulated mixture of silica, boron oxide, aluminium oxide, and trace additives that determined the glass's refractive index, dispersion characteristics, and transmission properties.
Optical glass was not a single material but a family of precisely engineered compositions. Each type had specific optical properties defined by two numbers: refractive index (how much the glass bent light) and Abbe number (how much it separated light into colours). A camera lens might use six different glass types in its construction, each chosen for specific optical correction. The slightest variation in composition — a fraction of a per cent difference in silica content — could shift the refractive index enough to make the glass useless for its intended application.
Manufacturing optical glass required melting raw materials at extreme temperatures, holding them molten for hours to achieve chemical homogeneity, then cooling them in precisely controlled ways to prevent internal stress. The entire process was part chemistry, part thermodynamics, part empirical art accumulated over centuries of glassmaking tradition.
Beside Karan stood Rajesh Trivedi, the facility's technical director — forty-three years old, trained as a materials scientist at IISc Bangalore, fifteen years of industrial experience, including four years at Schott Glass in Germany, before Karan had convinced him to return to India in early 1971 with an offer that was less about salary and more about the opportunity to build something that didn't exist.
"Furnace three is running BK7 today," Trivedi said, checking the control panel readings mounted on the gallery wall. "Borosilicate crown glass. Refractive index 1.5168, Abbe number 64.17. Standard optical glass for camera lenses, binoculars, and low-precision optics."
"Yield?" Karan asked.
"Eighty-three per cent on the last batch," Trivedi said. "Up from seventy-six per cent three months ago. We're still chasing the Germans — they get ninety per cent on BK7 — but the gap is closing."
Yield in optical glass manufacturing meant the percentage of molten glass that, after cooling and annealing, met the optical specifications for homogeneity, refractive index, and internal quality. The rest — glass with bubbles, stones (crystalline inclusions), or stress patterns — was scrapped and remelted. High yield meant efficient production and lower costs. Low yield meant expensive waste.
The difference between 76% and 83% yield on a production run of two tonnes meant an additional 140 kilograms of usable glass per batch — glass worth approximately ₹35,000 at current market prices for precision optical material.
"What changed?" Karan asked.
"Batch mixing precision," Trivedi said. "Ustad Ghulam was insisting the silica sand needed to be sieved differently. He said the standard 100-mesh screen was letting through particles that were creating seeds in the melt. We didn't believe him at first — the German specifications said 100-mesh was sufficient — but we tried 150-mesh on his recommendation. Yield jumped seven points immediately."
Karan looked at him. "Ustad Ghulam recommended this?"
"Yes," Trivedi said. "Based on forty years of melting glass by eye and hand. He couldn't explain the physics, but he knew what worked."
They walked along the gallery toward the annealing ovens, where cast glass cooled slowly over days to remove internal stress.
Firozabad had been India's glass capital for centuries — not optical glass, but bangles and decorative glass and utilitarian bottles, products made by families who had worked glass for generations using techniques passed down father to son in workshops that looked essentially unchanged since the Mughal era.
When Karan had decided to establish optical glass manufacturing in early 1971, the obvious locations were Bangalore (proximity to IISc and technical talent), Pune (existing industrial base) or Gorakhpur (integration with the semiconductor and aerospace facilities).
He had chosen Firozabad for a reason that looked eccentric on paper but made sense if you understood what optical glassmaking actually required.
The technical knowledge existed in textbooks and research papers — the chemical formulations, the melting temperatures, the annealing schedules. You could hire IISc-trained materials scientists who understood glass chemistry at the molecular level.
But there was another kind of knowledge that didn't exist in textbooks: the empirical understanding of how molten glass behaved, the subtle variations in viscosity that indicated composition shifts, the visual assessment of melt quality through the furnace viewport, the hand-feel of proper working temperature, the judgment calls that turned theoretical formulas into actual production.
That knowledge existed in Firozabad, accumulated over generations of glassmaking.
Karan's approach had been to combine them: hire the materials scientists for the theory, hire the traditional glassmakers for the practice, put them in the same facility, and let them teach each other.
It had worked better than expected.
The facility they walked through was a hybrid of traditional and modern.
The furnaces were modern — gas-fired regenerative furnaces with temperature control systems and atmosphere monitoring, designed to German specifications and built by Shergill Heavy Engineering to tolerances that traditional glassmakers had never imagined.
But operating the furnaces were men like Ustad Ghulam Mohammad — sixty-two years old, from a family that had made glass in Firozabad for six generations, a man who couldn't write a chemical formula but could tell you by looking at molten glass through a furnace port whether the batch was mixed correctly or needed adjustment.
"How many traditional glassmakers are working here now?" Karan asked as they descended the stairs toward the production floor.
"Forty-three," Trivedi said. "Plus sixty-two materials scientists, engineers, and technicians with formal education. The ratio has shifted over the past year. Early on, we needed more traditional expertise to get the processes working. Now that we're stable, we're hiring more engineers for expansion and new product development."
They reached the production floor.
The noise changed — the low roar of furnaces, the hiss of gas burners, the mechanical sounds of stirring mechanisms and temperature control systems, and underneath it all the particular silence of people working carefully around equipment hot enough to kill.
The main melting hall held six furnaces, each dedicated to a different glass type.
Furnace one: BK7 borosilicate crown (standard optical glass)
Furnace two: SF6 dense flint glass (high refractive index, for achromatic lens correction)
Furnace three: BK7 (redundant capacity, currently running)
Furnace four: Fused quartz (ultra-high purity for semiconductor applications)
Furnace five: Currently idle, scheduled for maintenance
Furnace six: Experimental melts (new formulations, small batches)
Each glass type required different raw materials, different melting temperatures, different holding times, and different annealing schedules. BK7 borosilicate was relatively forgiving — it melted at 1450°C and tolerated some composition variation. Dense flint glasses like SF6 were more demanding — they required lead oxide additions (making them heavier and higher refractive index) and precise melting to prevent lead reduction and discolouration. Fused quartz was the most extreme — pure silica melted at 1750°C in an electrically heated furnace, requiring platinum-lined crucibles to prevent contamination.
They stopped at furnace four, the quartz furnace.
Through the observation port, Karan could see the interior glowing white-hot, temperature readings showing 1720°C on the control panel.
"This is our strategic operation," Trivedi said quietly. "The quartz production."
He didn't need to explain why.
Fused quartz — ultra-high purity silica glass — was required for semiconductor manufacturing. The lithography equipment at ISMC used quartz lenses and windows. The diffusion furnaces used quartz tubes. The etching chambers used quartz components. Every advanced semiconductor process needed quartz that was transparent to ultraviolet light and chemically inert, and capable of withstanding thermal cycling without cracking.
Currently, semiconductor-grade quartz is imported from Heraeus in Germany or Corning in the United States at prices that reflect both the technical difficulty and the strategic importance.
Shergill Optical Materials had been producing semiconductor-grade quartz since September 1972 — six months after the facility opened. The first batches had quality issues (hydroxyl contamination causing UV absorption, crystalline inclusions, and dimensional instability). But by December, the problems were solved, and by February 1973, ISMC was running entirely on domestic quartz.
The cost savings were significant. But more important was the supply security — semiconductor manufacturing couldn't be held hostage to German export licenses or American trade restrictions.
"Current production rate?" Karan asked.
"Forty kilograms per week of semiconductor-grade quartz," Trivedi said. "ISMC's current consumption is about twenty-five kilograms per week. We're building inventory and preparing to supply other facilities when they come online."
"Quality specifications?"
"Hydroxyl content below 5 parts per million. Metal impurities below 0.1 parts per million. UV transmission at 254 nanometers is better than ninety-eight per cent. Thermal expansion coefficient within the specification range. We're meeting all the requirements."
Karan nodded. "What about scaling production?"
"We can double output with the current furnace by running continuous operation instead of batch mode. Beyond that, we'd need a second quartz furnace. That's a significant capital investment — the platinum lining alone costs eight lakh rupees."
"Do it," Karan said. "Order the second furnace. I want quartz production capacity ahead of semiconductor demand, not chasing it."
Trivedi made a note.
They walked toward the annealing area — a section of the facility where cast glass cooled slowly in computer-controlled ovens, the temperature dropping at rates measured in degrees per hour to prevent the internal stress that would ruin optical quality.
Annealing was critical because glass, despite being solid at room temperature, was technically a supercooled liquid — a disordered atomic structure frozen in place. When molten glass cooled too quickly, the outer surfaces solidified before the interior, creating permanent internal stress. This stress caused birefringence (double refraction), making the glass unsuitable for precision optics. Proper annealing required cooling the glass slowly enough that the entire piece remained in thermal equilibrium as it solidified — a process that took days for large pieces.
The annealing ovens held glass in various stages of cooling: thick disks destined to become telescope mirrors, optical blanks for camera lenses, rectangular slabs for laboratory windows.
Karan stopped at one oven showing a temperature of 380°C, well below the glass's softening point but still hot enough to be visibly glowing in the dim interior.
"What's in this one?" he asked.
"Telescope mirror blank," Trivedi said, checking the logbook mounted beside the oven. "Four hundred millimeters diameter, borosilicate, ordered by the Indian Institute of Astrophysics for a solar telescope they're building in Kodaikanal. Started annealing six days ago, another four days to go."
"We're supplying telescope mirrors now?"
"Small ones," Trivedi said. "The really large mirrors — meter-class and above — still require specialised facilities we don't have. But for research telescopes in the 200 to 500 millimetre range, we can produce blanks that meet astronomical specifications. The astrophysics community is very happy to have a domestic source."
They continued through the facility.
The grinding and polishing section, where optical blanks were shaped to precise curves using diamond grinding wheels and polishing compounds, the surfaces finished to tolerances measured in fractions of a wavelength of light.
The testing laboratory, where finished optics were measured for surface accuracy using interferometry, refractive index verified with refractometers, and dispersion characteristics checked against standard samples.
The quality control office, where every batch of glass was logged, tracked, and certified before shipment.
And throughout the facility, the combination of old and new: traditional glassmakers working alongside materials scientists, empirical knowledge informing theoretical understanding, centuries of craft integrated with modern precision.
They ended the tour in the development laboratory, a section of the facility dedicated to experimental work.
Dr Meera Kulkarni — thirty-four years old, PhD in materials science from IISc, specialist in glass chemistry — was running an experiment involving a small crucible furnace and a series of raw material samples arranged on the bench.
She looked up when Karan and Trivedi entered.
"Good timing," she said. "I was about to do the next melt test."
"What are you working on?" Karan asked.
"Laser glass," Kulkarni said. "Specifically, neodymium-doped glass for solid-state laser rods."
She gestured at the samples. "The principle is straightforward: add neodymium oxide to a suitable glass matrix — usually phosphate or silicate — and you get a material that can amplify light at 1064 nanometer wavelength when pumped with a flashlamp. It's used in everything from industrial cutting lasers to military rangefinders to scientific research."
"Currently imported?" Karan asked.
"Entirely," Kulkarni said. "From Schott, Hoya, Corning. Very expensive. A laser rod ten centimetres long and one centimeter diameter costs approximately one lakh rupees."
"And we can make it domestically?"
"I believe so," Kulkarni said. The glass chemistry is published in the literature. The challenge is achieving high optical quality with the neodymium dopant evenly distributed and no quenching centres that would kill the laser efficiency."
She loaded the crucible with a carefully measured mixture of chemicals — silica, phosphorus pentoxide, aluminium oxide, and a small amount of neodymium oxide that gave the powder a faint purple tint.
"This is my seventh attempt," she said, sliding the crucible into the furnace. "The first three cracked during cooling. The fourth had bubble inclusions. The fifth and sixth had acceptable quality but lower-than-expected laser efficiency. I'm adjusting the phosphate ratio and the annealing schedule."
The furnace ramped up to a temperature of 1400°C for this particular formulation.
"How long until you have a production-ready process?" Karan asked.
"Six months," Kulkarni said. "Maybe nine. Laser glass is more demanding than standard optical glass. But it's achievable."
Trivedi added, "Once we have laser glass production working, the market is significant. DRDO wants laser rangefinders. BARC wants high-power research lasers. Industrial cutting and welding applications are growing. Currently, everything is imported at high cost with long lead times."
"And strategically," Kulkarni said, "laser technology is going to matter enormously in the next two decades. Military applications, communications, materials processing, and medical surgery. Having domestic laser glass production positions us well."
Karan looked at the furnace where the experimental melt was proceeding.
"What else are you working on?" he asked.
Kulkarni pulled out a folder with several project sheets.
"Radiation-shielding glass for nuclear facilities — lead glass with high density and gamma-ray absorption. BARC is interested.
Fibre optic preforms — ultra-pure glass rods that can be drawn into optical fibres for telecommunications. This is early-stage, very difficult, but potentially transformative.
"Photochromic glass — glass that darkens in sunlight and clears indoors. Commercial application for eyeglasses and windows.
"Bioactive glass for medical implants — glass that bonds to bone tissue. Experimental but promising.
"Ultra-low expansion glass for telescope mirrors and precision instruments — essentially zero thermal expansion for applications requiring extreme dimensional stability."
She set down the folder. "We have more ideas than resources. The question is priorities."
"Laser glass first," Karan said. "That has immediate applications and clear demand. Fibre optic performs second — telecommunications will need it within five years. The others are longer-term research."
Trivedi made notes.
"There's one more thing," Kulkarni said. She looked at Karan directly. "We need to expand the traditional glassmaker training program."
"Explain," Karan said.
"When we started this facility, we hired experienced glassmakers from Firozabad. Men like Ustad Ghulam, who had decades of knowledge. That worked because we needed empirical expertise immediately. But those men are ageing. Ustad Ghulam is sixty-two. Mohammad Sharif is sixty-five. Ramesh Chandra is fifty-eight. In ten to fifteen years, that generation will retire."
"And?" Karan prompted.
"And their knowledge needs to be transferred to the next generation, but also documented and systematised," Kulkarni said. "We should establish a formal apprenticeship program. Take young people from Firozabad glassmaking families, teach them both the traditional craft and the modern science, document the empirical knowledge in usable form."
"We should also," she continued, "establish collaboration with glassmaking institutes internationally. Send our people to Schott in Germany, Corning in the US, and Hoya in Japan for advanced training. Bring experts here for short-term teaching. Build a knowledge network."
Trivedi nodded. "I agree with both points. The apprenticeship program preserves and expands the artisan knowledge base. The international collaboration keeps us connected to the cutting edge."
"How many apprentices are you proposing?" Karan asked.
"Twenty per year," Kulkarni said. "Starting immediately. Three-year program combining hands-on glassmaking with materials science fundamentals. By 1976, we had sixty trained people who understood both craft and science. By 1980, we had a hundred and twenty."
"That's expensive," Trivedi said. "Training costs, equipment, instructor time, and stipends for the apprentices."
"It's essential," Kulkarni countered. "Optical glass manufacturing isn't just running furnaces. It's an accumulated understanding of how glass behaves. If we lose that knowledge when the current generation retires, we'll have to relearn it expensively."
Karan considered this.
"Approved," he said. "Twenty apprentices per year, starting in April. Recruit from Firozabad primarily, but also from other traditional glassmaking centres — Moradabad, Purdalpur, wherever the craft knowledge exists. Document everything the master glassmakers know. Film the processes if necessary."
He paused.
"For the international collaboration — budget for sending six people per year abroad for three to six-month training stints. Also, budget for bringing two visiting experts per year to Firozabad for teaching residencies. I want this facility connected to the global glassmaking community, not isolated."
Kulkarni smiled. "That will make a significant difference."
They left the development laboratory and walked back toward the administrative section.
The facility was quieter here — offices, conference rooms, the quality assurance division, and the procurement department managing raw material supplies.
Trivedi's office overlooked the production floor through large windows, giving him a constant view of operations.
He sat down behind his desk and pulled out a planning document.
"You asked about expansion plans," he said. "I have a proposal."
He spread the document on the desk. It showed the current facility layout plus an adjacent plot of land marked for potential construction.
"Phase two expansion," Trivedi said. "Double the current furnace capacity, add specialised production lines for specific glass types, establish a dedicated R&D facility separate from production, and build additional annealing capacity."
"Cost?" Karan asked.
"One crore twenty lakh rupees over eighteen months. Construction, equipment, commissioning."
"Revenue justification?"
"Current revenue is approximately forty lakh rupees per month and growing by fifteen per cent quarterly. We're supply-constrained — we have orders we can't fill because furnace capacity is maxed out. The expansion would double production capacity, allowing us to serve growing demand from optics manufacturers, semiconductor facilities, scientific instruments, and emerging applications like laser systems."
Trivedi pulled out another sheet showing market projections.
"The Indian optics market is currently about twelve crore rupees annually, ninety per cent of which is imports. We're capturing about four per cent. With expanded capacity, we could reach fifteen per cent by 1975, twenty-five per cent by 1977."
"The semiconductor quartz market is smaller in rupees but strategically critical. As semiconductor manufacturing expands in India — and I know you're planning additional fabs — the quartz demand will grow proportionally. We should be ahead of that curve."
"The scientific instrument market — telescopes, microscopes, spectroscopy equipment, laboratory glassware — is maybe three crore rupees annually, all imports. We can capture a significant share."
"And emerging markets like fibre optics and laser systems could be ten to fifteen crore rupees by 1980, currently zero domestic production."
Karan studied the projections.
"The math works," he said. "But the expansion assumes we can maintain quality as we scale. Can we?"
"Yes," Trivedi said without hesitation. "The first eighteen months were learning and stabilisation. We now have documented processes, trained personnel, and proven quality systems. Scaling is about replicating what works, not inventing new processes."
"Then proceed," Karan said. "Authorise the expansion. Start design work immediately, construction by June."
Trivedi made a note. "I'll coordinate with the engineering division."
"One more thing," Karan said. "The facility needs a formal name. We've been calling it Shergill Optical Materials, which is functional but uninspiring. It should be something that reflects what this place actually is."
Trivedi thought for a moment.
"Bharat Optical Industries?" he suggested.
"Too generic," Karan said.
"Firozabad Precision Glass?"
"Doesn't capture the semiconductor and scientific applications."
They sat in silence for a moment.
Then Trivedi said: "What about simply: The National Glass Laboratory? It's accurate — this is a national-level capability in speciality glass. It's aspirational — laboratory implies research and development, not just production. And it positions us as a scientific institution rather than just a factory."
Karan considered this.
"National Glass Laboratory," he said. "NGL. I like it. It works for customers ranging from ISMC to telescope makers to laser research. It sounds serious."
"Then that's the name," Trivedi said.
The meeting concluded.
Karan walked back through the facility one more time before leaving, taking the long route through production to see the operations up close rather than from the observation gallery.
At furnace four, Ustad Ghulam Mohammad was supervising a quartz melt, watching through the observation port with the focused attention of someone reading a language that had no written form.
Karan stopped.
"Ustad-ji," he said. "How is the melt?"
Ghulam turned, recognised Karan, and gestured toward the port.
"See for yourself, sahib."
Karan looked through the port into the white-hot interior of the furnace.
The molten quartz — pure silica at 1720°C — was clear and colourless, viscous like honey, glowing from its own heat. Ghulam was right to watch it carefully. At this temperature, the slightest contamination would be visible as discolouration. The slightest temperature variation would affect viscosity and workability.
"It looks good," Karan said.
"It is good," Ghulam confirmed. "This will be a clean batch."
"How can you tell?" Karan asked.
Ghulam smiled slightly. "Forty-seven years of looking at glass, sahib. You learn to see what others don't see."
"Your recommendation about the finer sieving for the BK7 batches — Trivedi told me it improved yield significantly."
"The sand particles were too coarse," Ghulam said. "They don't melt completely at standard temperature. They become seeds in the glass. Finer sieving solves this."
"How did you know?"
"I have been melting glass since I was fifteen years old," Ghulam said. "My father taught me, his father taught him. You learn the feel of things. The scientists—" he gestured vaguely toward the offices "—they know the formulas. We know the glass."
"Both are necessary," Karan said.
"Yes," Ghulam agreed. "Together we make good glass. Separately—" he shrugged. "The scientists would make expensive mistakes. We would make beautiful glass that doesn't meet specifications. Together is better."
It was, Karan thought, an accurate summary of the facility's philosophy.
The combination of theoretical understanding and empirical mastery, modern precision and traditional craft, scientific measurement and artisan judgment.
Neither alone was sufficient. Together, they created a capability that matched anything produced internationally.
"One question, Ustad-ji," Karan said. "Your son — is he interested in learning the craft?"
Ghulam's expression shifted slightly. "My son is studying engineering at Roorkee. He wants to be a civil engineer, build bridges and dams. He thinks glassmaking is old-fashioned work for old-fashioned people."
"And what do you think?"
"I think he is young and will learn," Ghulam said. "But I also think the glassmaking families are becoming smaller. My generation, we all learned from our fathers. The next generation, some learn, some go to different work. The generation after that—" he shrugged.
"That's why we're starting the apprenticeship program," Karan said. "To preserve the knowledge and teach it to new people, whether they come from glassmaking families or not."
Ghulam looked at him. "You will pay people to learn glassmaking?"
"Yes," Karan said. "Stipend during training, salary after. We need the knowledge to continue."
"That is good," Ghulam said. "The knowledge should not die."
He turned back to the furnace port, checking the melt again.
"This batch will be ready to pour in ninety minutes," he said. "Clean quartz, good for the semiconductor work."
"Excellent," Karan said. "Keep up the good work, Ustad-ji."
He walked toward the exit.
The facility was producing forty kilograms of semiconductor-grade quartz per week, meeting the optical glass needs of camera manufacturers and telescope builders, developing laser glass for emerging applications, and training the next generation of glassmakers who understood both craft and science.
And it was all happening in Firozabad, a city known for bangles and decorative glass, now producing some of the most sophisticated optical materials in Asia.
The combination was unlikely but effective.
Sometimes the best solutions came from unexpected combinations.
Outside, the March afternoon was warm and hazy, Firozabad's industrial district visible in the distance with its traditional glass workshops sending smoke into the air from coal-fired furnaces that had operated essentially unchanged for generations.
Karan's car was waiting.
Before getting in, he looked back at the facility — a modern building, natural gas furnaces, precision equipment, and inside, the marriage of old knowledge and new science.
The National Glass Laboratory.
It had started as Shergill Optical Materials Division in 1971, a facility that didn't exist in any planning document until Karan decided it should exist.
Now it was a production facility, a research centre, a training ground, and a strategic asset that eliminated import dependence in a technology most people never thought about, but that everything else depended on.
Semiconductor manufacturing needed quartz. Cameras needed optical glass. Telescopes needed mirrors. Lasers needed specialised glass. Telecommunications would soon need fibre optics.
All of it required glass that was far more sophisticated than windows and bottles, glass that was engineered at the molecular level for specific optical properties, glass that combined chemical precision with artisan mastery.
And now India could make it.
Not importing it, not licensing it, not waiting for foreign suppliers to approve shipments.
Making it.
In Firozabad.
Where glassmakers had worked for centuries and would work for centuries more, except now they worked alongside materials scientists, and together they produced glass that competed with the best in the world.
Karan got in the car.
"Back to Gorakhpur," he told the driver.
As they pulled away, he made a note in his small notebook:
NGL expansion: approved. Apprenticeship program: 20/year starting in April. Laser glass: priority development. Fibre optic preforms: secondary priority.
Then he closed the notebook and looked out the window as Firozabad passed by — old and new, tradition and innovation, craft and science, all coexisting in the same small industrial city in Uttar Pradesh.
Sometimes that was how progress actually happened.
Not by abandoning the old but by integrating it with the new.
Not by importing solutions but by building capability from whatever resources existed, including resources that didn't look strategic until you understood what they actually were.
Firozabad had glassmaking knowledge accumulated over centuries.
Karan had given it modern science, precision equipment, market access, and strategic purpose.
Together they produced optical glass, semiconductor quartz, telescope mirrors, and soon laser rods.
The combination worked.
And it would keep working, expanding, improving, training new generations, developing new materials, serving new applications.
That was enough.
End of Chapter 111