LK-99 isn’t a superconductor — how science sleuths solved the mystery

Replications pieced together the puzzle of why the material displayed superconducting-like behaviours.

• Dan Garisto

Researchers seem to have solved the puzzle of LK-99. Scientific detective work has unearthed evidence that the material is not a superconductor, and clarified its actual properties.

The conclusion dashes hopes that LK-99 — a compound of copper, lead, phosphorus and oxygen — marked the discovery of the first superconductor that works at room temperature and ambient pressure. Instead, studies have shown that impurities in the material — in particular, copper sulfide — were responsible for the sharp drops in electrical resistivity and partial levitation over a magnet, which looked similar to properties exhibited by superconductors.

“I think things are pretty decisively settled at this point,” says Inna Vishik, a condensed-matter experimentalist at the University of California, Davis.

Claimed superconductor LK-99 is an online sensation — but replication efforts fall short

The LK-99 saga began in late July, when a team led by Sukbae Lee and Ji-Hoon Kim at the Quantum Energy Research Centre, a start-up firm in Seoul, published preprints1,2?claiming that LK-99 is a superconductor at normal pressure and temperatures up to at least 127 oC (400 kelvin). All previously confirmed superconductors function only at extreme temperatures and pressures.

The extraordinary claim quickly grabbed the attention of the science-interested public and researchers, some of whom tried to replicate LK-99.?Initial attempts did not see signs of room-temperature superconductivity, but were not conclusive. Now, after dozens of replication efforts, many experts are confidently saying that the evidence shows LK-99 is not a room-temperature superconductor. (Lee and Kim’s team did not respond to?Nature’s request for comment.)

Accumulating evidence

The South Korean team based its claim on two of LK-99’s properties: levitation above a magnet and abrupt drops in resistivity. But separate teams in Beijing, at Peking University3and the Chinese Academy of Sciences4?(CAS), found mundane explanations for these phenomena.

Another study5, by US and European researchers, combined experimental and theoretical evidence to demonstrate how LK-99’s structure made superconductivity infeasible. And other experimenters synthesized and studied pure samples6?of LK-99, erasing doubts about the material’s structure and confirming that it is not a superconductor, but an insulator.

The only further confirmation would come from the Korean team sharing their samples, says Michael Fuhrer, a physicist at Monash University in Melbourne, Australia. “The burden’s on them to convince everybody else,” he says.

Perhaps the most striking evidence for LK-99’s superconductivity was a?video?taken by the Korean team that showed a coin-shaped sample of silvery material wobbling over a magnet. The team said the sample was levitating because of the Meissner effect — a hallmark of superconductivity in which a material expels magnetic fields. Multiple unverified videos of LK-99 levitating subsequently circulated on social media, but none of the researchers who initially tried to replicate the findings observed any levitation.

Half-baked levitation

Several red flags popped out to Derrick van Gennep, a former condensed-matter researcher at Harvard University in Cambridge, Massachusetts, who now works in finance but was intrigued by LK-99. In the video, the same edge of the sample seemed to stick to the magnet, and it seemed delicately balanced. By contrast, superconductors that levitate over magnets can be spun and even held upside-down. “None of those behaviors look like what we see in the LK-99 videos,” van Gennep says.

He thought LK-99’s properties were more likely the result of ferromagnetism. So he constructed a pellet of compressed graphite shavings with iron filings glued to it. A?videomade by Van Gennep shows that his disc — made of non-superconducting, ferromagnetic materials — mimicked LK-99’s behaviour.

On 7 August, the Peking University team reported that this “half-levitation” appeared in their LK-99 samples because of ferromagnetism. “It’s exactly like an iron-filing experiment,” says Yuan Li, a condensed-matter physicist and study co-author. The pellet experiences a lifting force but it’s not enough to levitate — only enough to balance on one end.

Li and his colleagues measured their sample’s resistivity, and found no sign of superconductivity. But they couldn’t explain the sharp resistivity drop seen by the Korean team.

Impure samples

In their preprint, the Korean authors note one particular temperature at which LK-99’s showed a tenfold drop in resistivity, from about 0.02 ohm-centimetres to 0.002 ohm-cm. “They were very precise about it. 104.8oC,” says Prashant Jain, a chemist at the University of Illinois Urbana–Champaign. “I was like, wait a minute, I know this temperature.”

The reaction that synthesizes LK-99 uses an unbalanced recipe: for every 1 part copper-doped lead phosphate crystal — pure LK-99 — it makes, it produces 17 parts copper and 5 parts sulfur. These leftovers lead to numerous impurities — especially copper sulfide, which the Korean team reported in its sample.

Jain, a copper-sulfide expert, remembered 104oC as the temperature at which Cu2S undergoes a phase transition. Below that temperature, the resistivity of air-exposed Cu2S drops dramatically — a signal almost identical to LK-99’s purported superconducting phase transition. “I was almost in disbelief that they missed it.” Jain published a preprint7?on the important confounding effect.

On 8 August, the CAS team reported on the effects of Cu2S impurities in LK-99. “Different contents of Cu2S can be synthesized using different processes,” says Jianlin Luo, a CAS physicist. The researchers tested two samples — the first heated in a vacuum, which resulted in 5% Cu2S content, and the second in air, which gave 70% Cu2S content.

The first sample’s resistivity increased relatively smoothly as it cooled, and appeared similar to samples from other replication attempts. But the second sample’s resistivity plunged near 112 oC (385K) — closely matching the Korean team’s observations.

“That was the moment where I said, ‘Well, obviously, that’s what made them think this was a superconductor,’” says Fuhrer. “The nail in the coffin was this copper sulfide thing.”

Making conclusive statements about LK-99’s properties is difficult, because the material is finicky and samples contain varying impurities. “Even from our own growth, different batches will be slightly different,” says Li. But Li argues that samples that are close enough to the original are sufficient for checking whether LK-99 is a superconductor in ambient coniditions.

Crystal clear

With strong explanations for the resistivity drop and the half-levitation, many in the community were convinced that LK-99 was not a room-temperature superconductor. But mysteries lingered — namely, what were the material’s actual properties?

Initial theoretical attempts using an approach called density functional theory (DFT) to predict LK-99’s structure had hinted at interesting electronic signatures called ‘flat bands’. These are areas where the electrons move slowly and can be strongly correlated. In some cases, this behavior leads to superconductivity. But these calculations were based on unverified assumptions about LK-99’s structure.

To better understand the material, the US–European group5?performed precision X-ray imaging of their samples to calculate LK-99’s structure. Crucially, the imaging allowed them to make rigorous calculations that clarified the situation of the flat bands: they were not conducive to superconductivity. Instead, the flat bands in LK-99 came from strongly localized electrons, which cannot ‘hop’ in the way a superconductor requires.

On 14 August, a separate team, at the Max Planck Institute for Solid State Research in Stuttgart, Germany, reported6?synthesizing pure, single crystals of LK-99. Unlike previous synthesis attempts that relied on crucibles, the researchers used a technique called floating zone crystal growth that allowed them to avoid introducing sulfur into the reaction, eliminating the Cu2S impurities.

The result was a transparent purple crystal — pure LK-99, or Pb8.8Cu1.2P6O25. Separated from impurities, LK-99 is not a superconductor, but an insulator with a resistance in the millions of ohms — too high to run a standard conductivity test. It shows minor ferromagnetism and diamagnetism, but not enough for even partial levitation. “We therefore rule out the presence of superconductivity,” the team concluded.

The team suggests that the hints of superconductivity seen in LK-99 were attributable to Cu2S impurities, which are absent from their crystal. “This story is exactly showing why we need single crystals,” says Pascal Puphal, a specialist in crystal growth and the Max Planck physicist who led the study. “When we have single crystals, we can clearly study the intrinsic properties of a system.”

Lessons learned

Many researchers are reflecting on what they’ve learned from the summer’s superconductivity sensation.

For Leslie Schoop, a solid-state chemist at Princeton University in New Jersey, who co-authored the flat-bands study, the lesson about premature calculations is clear. “Even before LK-99, I have been giving talks about how you need to be careful with DFT, and now I have the best story ever for my next summer school,” she says.

Jain points to the importance of old, often overlooked data — the crucial measurements that he relied on for the resistivity of Cu2S were published in 1951.

While some commentators have pointed to the LK-99 saga as a model for reproducibility in science, others say that it’s an unusually swift resolution of a high-profile puzzle. “Often these things die this very slow death, where it’s just the rumors and nobody can reproduce it,” says Fuhrer.

When copper oxide superconductors were discovered in 1986, researchers leapt to probe their properties. But nearly four decades later, there is still debate over the material’s superconducting mechanism, says Vishik. Efforts to explain LK-99 came readily. “The detective work that wraps up all of the pieces of the original observation — I think that’s really fantastic,” she says. “And it’s relatively rare.”

doi: https://doi.org/10.1038/d41586-023-02585-7

References

1. Lee, S.?et al.?Preprint at arXiv?https://doi.org/10.48550/arXiv.2307.12037?(2023).

2. Lee, S.?et al.?Preprint at arXiv?https://doi.org/10.4855arXiv.2307.12008?(2023).

3. Guo, K., Li, Y. & Jia, S.?Sci.?China Phys. Mech. Astron. https://doi.org/10.1007/s11433-023-2201-9 (2023).

   Article?Google Scholar?

4. Zhu, S.?et al.?Preprint at arXiv?https://arxiv.org/abs/2308.04353?(2023).

5. Jiang, Y.?et al.?Preprint at arXiv?https://arxiv.org/abs/2308.05143?(2023).

6. Puphal, P.?et al.?Preprint at arXiv?https://arxiv.org/abs/2308.06256?(2023).

7. Jain, P. Preprint at arXiv?https://arxiv.org/abs/2308.05222?(2023).

中文翻譯

（由AI翻譯生成）

LK-99不是超導(dǎo)體——科學(xué)偵探如何解開這個(gè)謎團(tuán)

復(fù)制拼湊了為什么材料表現(xiàn)出超導(dǎo)樣行為的謎題。

• 丹·加里斯托

研究人員似乎已經(jīng)解決了LK-99的謎題?？茖W(xué)偵探工作發(fā)現(xiàn)了該材料不是超導(dǎo)體的證據(jù)，并澄清了其實(shí)際特性。

結(jié)論破滅了希望LK-99——一種由銅、鉛、磷和氧組成的化合物——標(biāo)志著第一個(gè)在室溫和環(huán)境壓力下工作的超導(dǎo)體的發(fā)現(xiàn)。相反，研究表明，材料中的雜質(zhì)——特別是硫化銅——是電阻率急劇下降和磁鐵上部分懸浮的原因，這看起來與超導(dǎo)體表現(xiàn)出的特性相似。

加州大學(xué)戴維斯分校的凝聚物質(zhì)實(shí)驗(yàn)家Inna Vishik說：“我認(rèn)為在這一點(diǎn)上，事情已經(jīng)非常果斷地解決了。”

聲稱超導(dǎo)體LK-99在網(wǎng)上引起了轟動(dòng)——但復(fù)制工作沒有達(dá)到要求

LK-99傳奇始于7月下旬，當(dāng)時(shí)首爾一家初創(chuàng)公司量子能源研究中心由Sukbae Lee和Ji-Hoon Kim領(lǐng)導(dǎo)的團(tuán)隊(duì)發(fā)表了預(yù)印本1,2，聲稱LK-99在正常壓力和溫度高達(dá)至少127oC（400開爾文）下的超導(dǎo)體。所有先前確認(rèn)的超導(dǎo)體僅在極端溫度和壓力下工作。

這一非同尋常的主張很快引起了對(duì)科學(xué)感興趣的公眾和研究人員的注意，其中一些人試圖復(fù)制LK-99。最初的嘗試沒有看到室溫超導(dǎo)的跡象，但并不確定?，F(xiàn)在，經(jīng)過數(shù)十次復(fù)制工作，許多專家自信地說，證據(jù)表明LK-99不是室溫超導(dǎo)體。（Lee和Kim的團(tuán)隊(duì)沒有回應(yīng)Nature的置評(píng)請(qǐng)求。）

積累證據(jù)

韓國(guó)團(tuán)隊(duì)的主張基于LK-99的兩個(gè)屬性：磁鐵上方的懸浮和電阻率的突然下降。但北京、北京大學(xué)3和中國(guó)科學(xué)院4（CAS）的獨(dú)立團(tuán)隊(duì)為這些現(xiàn)象找到了平凡的解釋。

美國(guó)和歐洲研究人員的另一項(xiàng)研究5結(jié)合了實(shí)驗(yàn)和理論證據(jù)，以證明LK-99的結(jié)構(gòu)如何使超導(dǎo)不可行。其他實(shí)驗(yàn)者合成并研究了LK-99的純樣品6，消除了對(duì)該材料結(jié)構(gòu)的懷疑，并確認(rèn)它不是超導(dǎo)體，而是絕緣體。

澳大利亞墨爾本莫納什大學(xué)的物理學(xué)家Michael Fuhrer說，唯一的進(jìn)一步確認(rèn)來自韓國(guó)團(tuán)隊(duì)分享他們的樣本。他說：“他們有責(zé)任說服其他人?！?/p>

也許LK-99超導(dǎo)性最引人注目的證據(jù)是韓國(guó)團(tuán)隊(duì)拍攝的一段視頻，該視頻顯示，一枚硬幣形狀的銀色材料樣本在磁鐵上搖晃。該團(tuán)隊(duì)表示，由于邁斯納效應(yīng)，樣品正在懸浮——這是材料排出磁場(chǎng)的超導(dǎo)性標(biāo)志。LK-99懸浮的多個(gè)未經(jīng)驗(yàn)證的視頻隨后在社交媒體上流傳，但最初試圖復(fù)制這些發(fā)現(xiàn)的研究人員都沒有觀察到任何懸浮。

半烤的懸浮

馬薩諸塞州劍橋哈佛大學(xué)前凝聚態(tài)研究員Derrick van Gennep發(fā)出了幾面紅旗，他現(xiàn)在從事金融工作，但對(duì)LK-99很感興趣。在視頻中，樣品的同一邊緣似乎粘在磁鐵上，似乎微妙地平衡。相比之下，懸浮在磁鐵上的超導(dǎo)體可以旋轉(zhuǎn)，甚至可以倒置。van Gennep說：“這些行為看起來都不像我們?cè)贚K-99視頻中看到的那樣?！?/p>

他認(rèn)為L(zhǎng)K-99的特性更可能是鐵磁性的結(jié)果。因此，他用鐵屑粘在上面制作了一塊壓縮石墨刨花顆粒。Van Gennep制作的一段視頻顯示，他的圓盤由非超導(dǎo)鐵磁性材料制成，模仿了LK-99的行為。

8月7日，北京大學(xué)團(tuán)隊(duì)報(bào)告說，由于鐵磁性，這種“半懸浮”出現(xiàn)在他們的LK-99樣本中。凝聚態(tài)物理學(xué)家、研究合著者袁立說：“這就像一個(gè)鐵歸檔實(shí)驗(yàn)?！睆椡枋艿教嵘?，但不足以懸浮——只夠在一端保持平衡。

李和他的同事測(cè)量了樣本的電阻率，沒有發(fā)現(xiàn)超導(dǎo)性的跡象。但他們無法解釋韓國(guó)隊(duì)看到的電阻率急劇下降。

不純的樣品

在他們的預(yù)印本中，韓國(guó)作者注意到一個(gè)特定溫度，LK-99的電阻率下降了十倍，從大約0.02歐姆-厘米下降到0.002歐姆-厘米?！八麄儗?duì)此非常精確。104.8oC，”伊利諾伊大學(xué)厄巴納-香檳分校的化學(xué)家Prashant Jain說。“我想，等一下，我知道這個(gè)溫度?！?/p>

合成LK-99的反應(yīng)使用不平衡的配方：每制造1份銅摻雜磷酸鉛晶體——純LK-99，它生產(chǎn)17份銅和5份硫。這些殘留物會(huì)導(dǎo)致許多雜質(zhì)——特別是硫化銅，韓國(guó)團(tuán)隊(duì)在其樣本中報(bào)告了這一點(diǎn)。

硫化銅專家Jain記得104oC是Cu2S發(fā)生相變的溫度。在該溫度以下，空氣暴露的Cu2S的電阻率急劇下降——這個(gè)信號(hào)幾乎與LK-99所謂的超導(dǎo)相變相同。“我?guī)缀醪桓蚁嘈潘麄冨e(cuò)過了?！盝ain發(fā)表了一份關(guān)于重要混淆效應(yīng)的預(yù)印本7。

8月8日，CAS團(tuán)隊(duì)報(bào)告了Cu2S雜質(zhì)在LK-99中的影響。CAS物理學(xué)家Jianlin Luo說：“Cu2S的不同內(nèi)容可以用不同的過程合成。”研究人員測(cè)試了兩個(gè)樣本——第一個(gè)在真空中加熱，產(chǎn)生了5%的Cu2S含量，第二個(gè)在空氣中產(chǎn)生了70%的Cu2S含量。

第一個(gè)樣品的電阻率隨著冷卻而相對(duì)平穩(wěn)地增加，并且看起來與其他復(fù)制嘗試的樣品相似。但第二個(gè)樣本的電阻率在112攝氏度（385K）附近暴跌——與韓國(guó)團(tuán)隊(duì)的觀察結(jié)果非常吻合。

“那是我說，'嗯，顯然，這就是讓他們認(rèn)為這是一個(gè)超導(dǎo)體'的時(shí)刻，”Fuhrer說?！肮撞睦锏尼斪邮沁@個(gè)硫化銅的東西?！?/p>

很難對(duì)LK-99的特性做出結(jié)論性陳述，因?yàn)椴牧虾芴籼蓿瑯悠分泻胁煌碾s質(zhì)。李說：“即使從我們自己的成長(zhǎng)來看，不同的批次也會(huì)略有不同。”但李認(rèn)為，與原件足夠接近的樣品足以檢查L(zhǎng)K-99在環(huán)境條件下是否是超導(dǎo)體。

晶瑩剔透

由于對(duì)電阻率下降和半懸浮的有力解釋，社區(qū)中的許多人確信LK-99不是室溫超導(dǎo)體。但謎團(tuán)揮之不去——即，材料的實(shí)際特性是什么？

使用一種稱為密度泛函理論（DFT）的方法來預(yù)測(cè)LK-99結(jié)構(gòu)的最初理論嘗試暗示了被稱為“扁平帶”的有趣電子簽名。這些是電子移動(dòng)緩慢且可以強(qiáng)烈相關(guān)的地方。在某些情況下，這種行為會(huì)導(dǎo)致超導(dǎo)性。但這些計(jì)算是基于對(duì)LK-99結(jié)構(gòu)的未經(jīng)驗(yàn)證的假設(shè)。

為了更好地了解材料，美歐組5對(duì)其樣本進(jìn)行了精確的X射線成像，以計(jì)算LK-99的結(jié)構(gòu)。至關(guān)重要的是，成像使他們能夠進(jìn)行嚴(yán)格的計(jì)算，以澄清扁平帶的情況：它們不利于超導(dǎo)性。相反，LK-99中的扁平帶來自強(qiáng)局部電子，這些電子不能像超導(dǎo)體要求的那樣“跳躍”。

8月14日，德國(guó)斯圖加特馬克斯·普朗克固態(tài)研究所的一個(gè)單獨(dú)團(tuán)隊(duì)報(bào)告了6合成LK-99的純單晶。與之前依賴坩堝的合成嘗試不同，研究人員使用了一種稱為浮區(qū)晶體生長(zhǎng)的技術(shù)，使其能夠避免將硫引入反應(yīng)中，從而消除Cu2S雜質(zhì)。

結(jié)果是透明的紫色晶體——純LK-99，或Pb8.8Cu1.2P6O25。與雜質(zhì)分離的LK-99不是超導(dǎo)體，而是電阻為數(shù)百萬歐姆的絕緣體——太高了，無法進(jìn)行標(biāo)準(zhǔn)的電導(dǎo)率測(cè)試。它表現(xiàn)出輕微的鐵磁性和變磁性，但甚至不足以進(jìn)行部分懸浮。“因此，我們排除了超導(dǎo)性的存在，”團(tuán)隊(duì)總結(jié)道。

該團(tuán)隊(duì)認(rèn)為，在LK-99中看到的超導(dǎo)性暗示可歸因于晶體中沒有的Cu2S雜質(zhì)。晶體生長(zhǎng)專家、領(lǐng)導(dǎo)這項(xiàng)研究的馬克斯·普朗克物理學(xué)家Pascal Puphal說：“這個(gè)故事恰恰說明了為什么我們需要單晶?！薄爱?dāng)我們有單晶時(shí)，我們可以清楚地研究一個(gè)系統(tǒng)的內(nèi)在屬性?！?/p>

吸取的教訓(xùn)

許多研究人員正在反思他們從夏天的超導(dǎo)感覺中學(xué)到的東西。

對(duì)于新澤西州普林斯頓大學(xué)固態(tài)化學(xué)家Leslie Schoop來說，他共同撰寫了扁平帶研究，關(guān)于過早計(jì)算的教訓(xùn)是明確的。她說：“甚至在LK-99之前，我一直在談?wù)撃阈枰绾涡⌒腄FT，現(xiàn)在我的下一個(gè)暑期學(xué)校有了有史以來最好的故事?！?/p>

Jain指出了經(jīng)常被忽視的舊數(shù)據(jù)的重要性——他依賴的Cu2S電阻率的關(guān)鍵測(cè)量結(jié)果于1951年發(fā)布。

雖然一些評(píng)論家指出LK-99傳奇是科學(xué)中可重現(xiàn)性的模型，但其他人表示，這是一個(gè)高調(diào)謎題的異?？焖俚慕鉀Q。Fuhrer說：“通常這些東西會(huì)死得非常緩慢，這只是謠言，沒有人可以重現(xiàn)它。”

當(dāng)1986年發(fā)現(xiàn)氧化銅超導(dǎo)體時(shí)，研究人員跳來探索它們的特性。但近四十年后，關(guān)于這種材料的超導(dǎo)機(jī)制仍然存在爭(zhēng)議，Vishik說。解釋LK-99的努力是很容易的。她說：“偵探工作總結(jié)了原始觀察的所有部分——我認(rèn)為這真的很棒。”“而且這相對(duì)罕見?！?/p>

doi：https://doi.org/10.1038/d41586-023-02585-7

參考文獻(xiàn)

1. Lee，S.等人。在arXiv?https://doi.org/10.48550/arXiv.2307.12037?(2023)預(yù)印本。

2. Lee，S.等人。預(yù)印本在arXiv?https://doi.org/10.4855arXiv.2307.12008?(2023)。

3. 郭，K.，李，Y。& Jia, S.科學(xué)。中國(guó)物理機(jī)甲。Astron。https://doi.org/10.1007/s11433-023-2201-9 (2023)。

   文章?谷歌學(xué)術(shù)?

4. Zhu，S.等人。在arXiv?https://arxiv.org/abs/2308.04353?(2023)預(yù)印本。

5. Jiang，Y.等人。arXiv?https://arxiv.org/abs/2308.05143?(2023)上的預(yù)印本。

6. Puphal，P.等人。arXiv?https://arxiv.org/abs/2308.06256?(2023)上的預(yù)印本。

7. Jain，P。在arXiv?https://arxiv.org/abs/2308.05222?(2023)預(yù)印本。

原文：https://www.nature.com/articles/d41586-023-02585-7